Digital Infrastructure - Subject Guide

Undergraduate study in Economics,
Management, Finance and the Social Sciences
Digital
infrastructures
for business
M. Hercheui
T. Cornford
IS1181
2022
Digital infrastructures
for business
M. Hercheui
T. Cornford
IS1181
2022
Undergraduate study in
Economics, Management,
Finance and the Social Sciences
This subject guide is for a 100 course offered as part of the University of London
undergraduate study in Economics, Management, Finance and the Social
Sciences. This is equivalent to Level 4 within the Framework for Higher Education
Qualifications in England, Wales and Northern Ireland (FHEQ).
For more information, see: london.ac.uk
This guide was prepared for the University of London by:
Dr Magda Hercheui, UCL School of Management, University College London.
Dr Tony Cornford, Department of Management, London School of Economics and
Political Science.
This is one of a series of subject guides published by the University. We regret that
due to pressure of work the authors are unable to enter into any correspondence
relating to, or arising from, the guide. If you have any comments on this subject
guide, favourable or unfavourable, please use the form at the back of this guide.
University of London
Publications Office
Stewart House
32 Russell Square
London WC1B 5DN
United Kingdom
london.ac.uk
Published by: University of London
© University of London 2022
The University of London asserts copyright over all material in this subject guide
except where otherwise indicated. All rights reserved. No part of this work may be
reproduced in any form, or by any means, without permission in writing from the
publisher. We make every effort to respect copyright. If you think we have inadvert-
ently used your copyright material, please let us know.
Contents
Contents
Chapter 1: Introduction to the course..................................................................... 1

1.1 Route map to the guide............................................................................................ 1
1.2 Introduction to the subject area of digital infrastructures for business ....................... 2
1.3 Syllabus ................................................................................................................... 4
1.4 Aims of the course ................................................................................................... 5
1.5 Learning outcomes for the course ............................................................................ 5
1.6 Overview of learning resources ................................................................................ 5
Chapter 2: Infrastructures: conceptual beginnings............................................... 11
Chapter outline............................................................................................................ 11
2.1 Introduction........................................................................................................... 11
2.2 Chapter content .................................................................................................... 13
2.3 Overview of chapter .............................................................................................. 31
2.4 Reminder of learning outcomes ............................................................................. 32
2.5 Test your knowledge and understanding ................................................................ 32
Chapter 3: History of the internet ........................................................................ 33
Chapter outline............................................................................................................ 33
3.1 Introduction........................................................................................................... 33
3.2 Chapter content..................................................................................................... 34
Chapter 4: Emergent design of the internet ........................................................ 45
Chapter outline............................................................................................................ 45
4.1 Introduction........................................................................................................... 45
4.2 Chapter content .................................................................................................... 46
Chapter 5: The internet today and the TCP/IP ...................................................... 53
Chapter outline............................................................................................................ 53
5.1 Introduction........................................................................................................... 53
5.2 Chapter content .................................................................................................... 54
Chapter 6: Using a layered model of infrastructure: ‘above and below’.............. 65
Chapter outline............................................................................................................ 65
6.1 Introduction........................................................................................................... 65
6.2 Chapter content .................................................................................................... 66
i
IS1181 Digital infrastructures for business
Chapter 7: Core ideas for evolving infrastructures I: installed base

and cultivation...................................................................................................... 77
Chapter outline............................................................................................................ 77
7.1 Introduction........................................................................................................... 77
7.2 Chapter content..................................................................................................... 79
Chapter 8: Core ideas for evolving infrastructures II: standards and
network externalities............................................................................................ 87
Chapter outline............................................................................................................ 87
8.1 Introduction........................................................................................................... 87
8.2 Chapter content .................................................................................................... 88
8.3 Overview of chapter............................................................................................... 98
Chapter 9: Economics and institutional perspectives on
digital infrastructures............................................................................................ 99
Chapter outline............................................................................................................ 99
9.1 Introduction........................................................................................................... 99
9.2 Chapter content .................................................................................................. 101
9.3 Overview of chapter ............................................................................................ 110
9.4 Reminder of learning outcomes ........................................................................... 110
9.5 Test your knowledge and understanding .............................................................. 110
Chapter 10: The world wide web and web engineering..................................... 113
Chapter outline.......................................................................................................... 113
10.1 Introduction....................................................................................................... 113
10.2 Chapter content ................................................................................................ 115
10.3 Overview of chapter .......................................................................................... 120
10.4 Reminder of learning outcomes ......................................................................... 120
10.5 Test your knowledge and understanding ............................................................ 121
Chapter 11: Cloud computing as an innovation in infrastructure....................... 123
Chapter outline.......................................................................................................... 123
11.1 Introduction....................................................................................................... 123
11.2 Chapter content ................................................................................................ 125
Chapter 12: Mobile technology as an innovation in infrastructure.................... 133
Chapter outline.......................................................................................................... 133
12.1 Introduction....................................................................................................... 133
12.2 Chapter content ................................................................................................ 136
12.3 Overview of chapter........................................................................................... 143
ii
Contents
Chapter 13: Social media as an innovation in infrastructure.............................. 145

Chapter outline.......................................................................................................... 145
13.1 Introduction....................................................................................................... 145
13.2 Chapter content ................................................................................................ 147
Chapter 14: Delivering digital products and digital services.............................. 157
Chapter outline.......................................................................................................... 157
14.1 Introduction....................................................................................................... 157
14.2 Chapter content ................................................................................................ 158
Chapter 15: Governance and regulation of digital infrastructures..................... 171
Chapter outline.......................................................................................................... 171
15.1 Introduction....................................................................................................... 171
15.2 Chapter content ................................................................................................ 173
Chapter 16: Issues of net neutrality and national borders................................. 183
Chapter outline.......................................................................................................... 183
16.1 Introduction....................................................................................................... 183
16.2 Chapter content ................................................................................................ 185
Chapter 17: Digital business platforms and digital infrastructures.................... 193
Chapter outline.......................................................................................................... 193
17.1 Introduction....................................................................................................... 193
17.2 Chapter content ................................................................................................ 195
Chapter 18: Business models and digital infrastructures.................................... 205
Chapter outline.......................................................................................................... 205
18.1 Introduction....................................................................................................... 205
18.2 Chapter content ................................................................................................ 206
Chapter 19: Smart cities...................................................................................... 217
Chapter outline.......................................................................................................... 217
19.1 Introduction....................................................................................................... 217
19.2 Chapter content ................................................................................................ 218
iii
Chapter 20: Reflections on the course: the future of digital infrastructures

for business......................................................................................................... 229
Chapter outline.......................................................................................................... 229
20.1 Introduction....................................................................................................... 229
20.2 Chapter content................................................................................................. 231
iv
Chapter 1: Introduction to the course
1.1 Route map to the guide

Welcome to the subject guide for IS1181 Digital infrastructures for
business.
In this first chapter, you will gain an overall view of the subject you
are going to study in the next months, detailing the syllabus, aims and
learning outcomes of the course. We also introduce to you the learning
resources you have available for succeeding in IS1181. These start
with this subject guide and then include specified essential and further
readings, online resources including videos and the various resources
found in the VLE (virtual learning environment). At the end of the chapter,
we discuss the examination and give some advice on how to ensure that
you are well prepared when the exam day comes.
Before we introduce the content of the course, we want to give you an
overview of the subject area and the approach you need to take to your
study for you to reach your learning objectives. Indeed, the foremost
purpose of this guide is to help you to find the best ways to approach
your studies and to show you how the various parts of the syllabus link
together. Of course, it is for you then to do the work and that means
studying using several resources to find relevant information and
contrasting opinions. To access this information, you will be working with
textbooks, materials in the VLE and other online resources. Each chapter
contains quite detailed reading guidance as well as assignments and
activities to help you expand and consolidate your understanding. It is
very important that you undertake all these assignments; they are central
to your learning. The assumption we make when setting exam papers
is that you have done all of them and thus can report on the specific
experience in exam questions.
Your study should not be limited by the materials and resources directly
indicated in this guide. The more you read and the more you discuss, the
deeper your knowledge will become. This has two linked benefits. First,
any subject becomes more interesting as you ‘get deeper into it’. Second,
deeper knowledge combined with good exam technique, equals better
exam results.
This course should help you in a wider sense to understand digital changes
occurring around you and perhaps to find new employment opportunities
in a world that is more and more reliant on digital infrastructures. As you
can see around you, digital technology using new networks and offering
new services has become pervasive in almost all societies, industries,
sectors and economies. Not long ago, knowing how to use the Microsoft
Office package and the internet would be enough for a graduate student
to enter a good career in government or business organisations. Not
anymore! As the digitalisation of the economy progresses, business
professionals are more and more required to have up-to-date knowledge
and understanding of digital infrastructures and more importantly the
kinds of change they bring – for example all manner of new products and
services, new business models and new ways of organising. One author,
reflecting on these kinds of change, has defined digitalisation as,
1
‘The encoding of analogue information into a digital format and

the possible subsequent reconfigurations of the socio-technical
context of production and consumption of the associated
products and services’ (Yoo 2013).
In this spirit of ‘consequential change’, we will consider in this course

the potential for innovation that digital infrastructures bring – where
innovation may be seen in new products, new service, changes on
organisational and market structures and perhaps in new attitudes and
behaviours. The kind of knowledge that understands such change can also
offer opportunities for entrepreneurs who want to create new businesses
and serve new markets, meeting new needs in society and across the
global economy. So, one objective of this course is to help you to gain
the knowledge you need to be positioned for success in the global digital
economy.
To reach this objective, this subject guide aims to:
• present relevant subject material in a structured and clear format,
organised into sections and chapters, allowing you to gradually
advance in the understanding of a complex topic
• guide you to use the available learning resources and motivate you to
look for further sources in accordance with your developing interests
• encourage you to take an active learning approach, through reading
the materials, undertaking learning activities and engaging in
discussion with peers via the VLE.
In the following sections, we present how this subject guide is going to
deliver these three aims.
1.2 Introduction to the subject area of digital

infrastructures for business
This guide is structured to offer you a smooth, structured learning path.
We have divided the subject guide into 20 chapters. This first chapter
is the introduction to the guide, discussing the overall structure of the
course and its assessments. In the following 18 chapters we introduce and
develop in depth key aspects of digital infrastructures for business. The last
chapter summarises the key ideas introduced, the discussions opened up
and some key issues for the future development of digital infrastructures.
The concept of digital infrastructures is established in Chapter 2.
Summarising the key ideas behind this concept, we can say that digital
infrastructures are fundamental digital systems and services which serve as
the base for many distinct and separate activities. The most fundamental
of such services is the communication of data across digital networks
(e.g. the internet), but digital infrastructures extend beyond just data
transmission. They include many kinds of enablers of a particular way
of doing things and meeting needs – a digital way that exploits basic
data transmission. Examples might include social media platforms such
as Facebook or LinkedIn, digital marketplaces such as Alibaba, or cloud
services as offered by Microsoft, IBM, SalesForce or Amazon Web Services
(AWS). We also include a wider range of digital business platforms (DBP)
that support various industry sectors such as travel, finance, e-commerce,
media or entertainment. All such infrastructure will be embodied in and
dependent on, physical, network and organisational resources and they
will be shared by a large and diverse set of people and organisations.
2
In this way the most significant and pervasive digital infrastructure,

the internet, should be seen as a combination of hardware, software,
agreed standards, operating practices as well as institutions to coordinate
activity and rules of governance. The internet provides an infrastructure
(a communication infrastructure) for businesses, governments and
individuals. It is used in a huge variety of ways according to the needs of
different people or organisations. In a similar way, the world wide web
(WWW) is equally heterogeneous (made of different kinds of things) and
provides an information infrastructure.
In Chapter 2 we give more details on how the concept of an infrastructure
has developed and we introduce the four principal digital infrastructures
that are discussed in this subject guide and which themselves are built
on top of the internet, 1) the WWW, 2) cloud computing, 3) mobile
technologies and 4) social media. We also introduce the concept of
innovation, which is fundamental for our understanding of the relevance
of digital infrastructures.
In the next three chapters (3–5), we turn first to more technical issues
as we consider the internet. These chapters bring fundamental technical
knowledge that all business and management professionals should have
in order to better understand and use digital infrastructures. Chapter 3
explains the history of the internet. Chapter 4 considers how the internet
design emerged through time to become our most fundamental digital
infrastructure. In Chapter 5 the TCP/IP (rules of operations) at the heart
of the internet are explained.
Chapters 6 to 9 develop conceptual frameworks (models and theories)
necessary to understand and analyse digital infrastructures. Chapter 6
introduces the key conceptual framework used in this subject guide: the
layered model, also known as the hourglass architecture. Chapters
7 and 8 develop further theoretical ideas on how digital infrastructures
evolve, focusing on issues of installed base, cultivation, standards, lock-
in and network externalities. An example of Application Programming
Interface (API) is explored, demonstrating the significance of standards
and how they support the idea of cultivating the installed base and
enabling network externalities. These concepts are used in subsequent
chapters when discussing specific examples of digital infrastructures.
Chapter 9 expands the discussion in another direction, exploring economic
and institutional perspectives on digital infrastructures and the expansion
of digital infrastructures from a macro perspective.
After discussing theoretical perspectives, the next five chapters (10–14)
show how the internet supports other important digital infrastructures,
which themselves support a myriad of digital products and services.
Chapter 10 moves on from the internet to consider the WWW and web
engineering. Chapters 11, 12 and 13 present in sequence cloud computing,
mobile technologies and social media, each seen as digital infrastructures.
Chapter 14 brings the discussion to consider in a more general way digital
products and services, exploring how they are connected to or enabled by
digital infrastructures. These chapters consider digital infrastructures more
from the perspective of businesses and apply the models and concepts
discussed in previous chapters (e.g. Chapter 6).
Chapters 15 and 16 cover the governance and regulation of digital
infrastructures. Chapter 15 explains the concepts of governance and
regulations in the sphere of the internet and other digital infrastructures.
It also introduces key internet governance institutions. Chapter 16 extends
this discussion on governance to two specific contentious issues, net
3
neutrality and national borders. These issues illustrate how governance

and regulation of digital infrastructures may be controversial. These
chapters also emphasise that digital infrastructures are not just technical
systems (hardware and software), but are a sociotechnical system that
depends on technology and people and agreed practices, norms, rules and
regulations as much as hardware and software.
Chapters 17 and 18 explore the broader implications of digital
infrastructures in enabling change in organisations and markets, first in
the development of digital business platforms and second the innovation
of business models. Both chapters bring theoretical perspectives which are
applied to these topics. Chapter 19 presents a detailed example of how
digital infrastructures may affect our lives in a broad sense through the
example of smart cities.
Finally, Chapter 20 summarises the key ideas introduced in the course,
explores some implications of the way digital infrastructures are evolving.
The rate of change in the digital business domain is so great that it is
exceedingly difficult to forecast the next big events and challenges.
However, we may have some clues from more recent development of
technologies, in areas such as artificial intelligence, algorithms and big
data analytics. We hope the last chapter motivates you to take other
related courses within your degree programme and to appreciate the
relevance of learning how to learn.
1.3 Syllabus
The course provides an introductory understanding of the concept of an
infrastructure with a particular focus is on the digital infrastructures that
underpin so much of business, government and social life – including the
internet and world wide web and the many other systems and services
that build upon these. Concepts and theories are introduced that help
understand the emergence of digital infrastructures, their architecture
(structure), their economics and the kinds of services they offer including
access to data services and communications capacity.
Students learn to apply a critical perspective to this topic, developing
understanding in how to exploring the opportunities and risks of
these infrastructures as seen by various groups of people. The course
considers the issues and challenges faced by decision makers in
government, businesses and infrastructure provider organisations as
digital infrastructures become more central in economic and social life.
This includes issues of national and international standards, laws and
regulation. The course also addresses these issues as they relate to the
individual user of digital services.
To provide focus and depth, the course focuses on the internet as our most
general and fundamental digital infrastructure for communication and the
world wide web as the core information infrastructure. It also considers
three other digital infrastructures of relevance today: cloud computing,
mobile devices and services and social media. You will learn the key
characteristics of the technologies that underpin each case, how these
infrastructures have emerged and how they are being used by individuals,
businesses and governments. The course allows students to obtain a robust
understanding of the theoretical underpinnings of current developments
in digital infrastructures and to apply this to other contemporary examples
including concepts of platforms, business model innovation and the
Internet of Things (IoT).
4
1.4 Aims of the course

The course introduces the concepts of digital infrastructures and their
core technologies. It considers how digital infrastructures are established
and evolve, their design, economics and how they support innovation
and business change. It introduces models of infrastructure development
including concepts of the installed base, cultivation and path dependency.
The course considers the internet and the WWW and innovations that
exploit digital infrastructures, including cloud computing, mobile services
and social media. The course considers the opportunities and risks that
digital infrastructures present, including a consideration of regulatory
issues.
1.5 Learning outcomes for the course

At the end of the course and having completed the Essential readings and
activities, you should be able to:
• describe the concept of an infrastructure and relate it to digital systems
and services
• discuss how digital infrastructures are structured and how they are
used
• identify and apply theories and models to understand the evolution of
digital infrastructures and their roles in fostering innovation
• analyse three contemporary examples of digital infrastructures in use:
cloud computing, mobile services and social media, using relevant
examples
• understand how and why some digital infrastructures become globally
pervasive
• assess relevant issues of public policy, laws and regulations in regional,
national and international settings
• apply concepts and theories learnt to other emergent digital
infrastructures.
1.6 Overview of learning resources

You have access to resources in the University of London’s Online Library,
in addition to the local resources of your university library. This guide also
refers extensively to other online resources, including Wikipedia content,
institutional websites, white papers, consultancy reports and videos.
We make particular use of Wikipedia as a learning resource in this course.
Wikipedia is often not regarded very highly in education – often seen as
a dubious source of information and an invitation to learn things in small
‘sound bites’. For the material we cover in this course there are, we believe,
very positive reasons to use Wikipedia and other online sources. After all
this is a course on digital infrastructures and what better way to learn
about them than to use them? It is also the case that the community that
builds and develops digital infrastructures is a community that expects
to keep its records online. Put more bluntly, the way Wikipedia works,
drawing on volunteer experts, is perfectly adapted to knowledge about
technical and digital projects. While it may not serve for a study of history,
or geography, of politics or economics, it certainly does serve for any study
in the digital domain. This is not to say that everything online is true, good,
or useful – far from it – but a source like Wikipedia, with an open and well-
managed system of governance for its materials, is valid and useful.
5
The subject guide also refers to books. When these books are not available
in the local library or online, you can try to find the referred chapters
online on Google Books. Some authors make electronic copies of their
books available. Authors also may have online papers which discuss similar
topics as the books. Finally, you can use the virtual learning environment
(VLE) resources, as discussed below in section 1.6.4. For those who like
learning from videos, we recommend you search for experts in the fields of
your interest, to learn from a variety of professionals.
1.6.1 The subject guide

This subject guide provides a summary of the topics and theories you are
going to learn in this course. It is intrinsic to your studies.
To get the best from this guide, its chapters should be read before the
other resources and in preparation for each lecture if you are studying in
a formal institution. Before a lecture, read the subject guide to understand
the key concepts and theories to be introduced and developed and prepare
key questions you are going to ask during the lecture. For students
studying on their own we similarly recommend that you read a chapter
at the start of a topic, then go on to further reading and activities and
then return to the chapter at the end to cross check your knowledge. In
general, we expect that at the first reading chapters will present ideas that
you are unfamiliar with and may find difficult, but that when you return
after active study using other resources, the chapter will make more sense
and even seem a bit too simplistic! In any case, it is important that you
engage actively in your learning, in and work to overcome the difficulties
of getting knowledge on something new.
After the lecture, if you still have doubts on some parts, read the subject
guide again, highlighting key ideas and mapping your own understanding
on the topic. Organise your material since the beginning to facilitate your
exam revision.
With these key concepts and theories in mind, you can better explore
further readings of the core texts and beyond. Each chapter in this guide
brings a list of core texts and additional readings, guiding you also in the
selection of these additional resources.
All the topics related to digital infrastructures for business are complex
with a variety of ideas to be explored. The guide then provides you
with a road map to help you to focus on the key aspects. This is an
introductory course and we do not expect enormous depth of knowledge
so one purpose of this guide is to help students to understand the level
of knowledge expected from them to succeed in the course. And in
many areas, we have had to choose specific topics and ideas to study
from among a larger potential pool. But we do not want to limit your
enthusiasm for knowledge! So, the guide does provide ideas for further
readings for students who go further. In general, we would say that all
students should do some of the further reading on topics they choose, but
that we cannot expect all students to do all such reading.
To facilitate the process of learning, the guide lists the essential readings
for each chapter and some further readings. All these essential readings
are accessible to the students via the VLE for those who want to add depth
to their understanding of the syllabus. You need to read systematically
from the first week of the course. Do not delay these activities in any
circumstance. Remember the syllabus of each lecture draws upon your
knowledge from the previous lectures. If you delay your readings and
preparation, chances are that you are not going to fully understand
6
subsequent lectures and chapters, reducing your chances of getting the

best from this module.
1.6.2 Essential reading

We are unable to specify a single textbook as essential reading for this
course because, so far as we know, none is published (yet) covering the
basic ideas of digital infrastructures and their uses. This is not surprising
because until recently information systems degrees would not have
contained a course quite like this and if they did it would have been a
second- or third-year option most probably focused on ‘networks and
networking’. This is not to say that the topics we study here are not
written about – far from it as you will see. Our view is, however, that
times have changed and today in the third decade of the 21st century
understanding digital infrastructures and the internet is foundational for
anybody studying information systems and digital innovation.
We are, however, lucky because this lack of a single book is made less of
a problem by the availability of digital infrastructures! That is, there is
plenty of good study material available on the internet in magazines and
journals as well as in video presentations. This is the route we take in this
guide and we have devoted a good deal of time and effort to pull together
a suitable set of resources to support your study. These resources will be
updated from time to time on the VLE and you should certainly look there
for up-to-date links and not assume that this printed guide still contains
the latest materials when you come to read it perhaps a few years from the
guide’s preparation.
There are, however, a few books that we do recommend that cover some
of the course syllabus. The first book we recommend is Laudon and
Laudon (2021), or other similar editions of the same book.
Laudon, K.C. and J.P. Laudon Management information systems – managing
the digital firm. (Harlow, England: Pearson Education, 2021) 17th edition
[ISBN 9781292417752].
You can use any of the last or later editions of this book, as the authors
update it very regularly, but the essence of the discussion is very similar
(with bigger changes in the examples used). You just need to look for the
sections in the book which discuss the topics of each of the chapters in this
subject guide.
This book is also the recommended course text for the companion
course IS1060 Introduction to information systems. Students taking
the full Digital Innovation and Management Degree must take IS1060
Introduction to information systems as well as IS1181. It is thus a good
idea to make use of this text as much as possible for both courses, but it is
also important to explain a bit about how the two courses differ.
In simple terms IS1060 Introduction to information systems
and the book recommended there (Laudon and Laudon, 2021) take the
perspective of an organisation or business that needs information systems
to serve their own needs, systems that are in part built using information
and communications technologies and that are then embedded in the
organisation and its ways of working. That is, the perspective of an
organisation that builds and uses their own business information systems.
The organisations considered in IS1060 may be of various types, a charity,
a small business, a multinational corporation, a government department
etc., but they each will have managers and other kinds of technical
professionals (project managers, programmers, systems analysts, database
administrators) whose job is to plan, procure, configure and run these
7
information systems. We can also usually assume that the organisation

gains (or wants to gain) benefit from developing and using their systems,
for example in terms of efficiency, competitive advantage or innovation.
In contrast, this course IS1181 Digital infrastructures for business
takes the perspective of the digital technologies and services that these
organisations (and their people) draw upon in order to develop and
run information systems – we call these digital infrastructures. The
emphasis is not on the organisation, its managers and the users (or uses)
of their information systems. Rather, the focus is on the kinds of digital
communications services used (e.g. the internet, mobile phones etc.)
and the services these enable for business information systems. ‘Digital
infrastructures’ is the collective name we use for these services, including
services for data transmission and communications from place to place
(e.g. the internet, mobile phones), information provision (e.g. the world
wide web) and social technologies and collaboration (e.g. Facebook).
But it is not just about or for business and formal organisations. Individual
people also use digital infrastructures, so we are concerned with the
needs of you and me as well. We are all users of digital infrastructures, for
example, the internet, WWW and social media like Instagram, WeChat or
Facebook. We also as individuals use e-commerce services such as Amazon,
eBay or Alibaba, read newspapers online or stream films and television
shows. So, digital infrastructures are what links us all to each other
and links us to commercial businesses, not-for-profit organisations and
government bodies (and to the University of London too).
Another, much simpler, way to make the distinction between the two
courses is to say that IS1060 is about ‘inside’ the organisation, while
IS1181 is about the digital environment ‘outside’ and interlinking many
organisations and individuals of all sorts.
The second book we recommend is Zittrain (2009):
Zittrain, J.L. The future of the internet – and how to stop it. (New Haven and
London: Yale University Press, 2009) [ISBN 9780300151244] Chapter 4
‘The generative pattern’ pp.67–100. You may find a copy of this book online
through Google Search or Google Scholar.
1.6.3 Further reading

Brynjolfsson, E. and A. McAfee The second machine age. (New York, London:
W.W. Norton, 2014) [ISBN 9780393350647].
In so far as digital infrastructures underpin the digital economy, we
recommend Brynjolfsson and McAfee (2014) as useful introductory
reading. We refer to a few chapters as specific further reading in the later
parts of this guide.
Constantinides, P. Perspectives and implications for the development
of information infrastructures. (Cyprus: IGI Global, 2012)
[ISBN 9781466616226].
Kernighan, B. Understanding the digital world: what you need to know about
computers, the internet, privacy and security. (New Jersey: Princeton
University Press, 2017) [ISBN 9780691176543]. Also available as an
ebook. See in particular section 3 on networks and infrastructures.
Kitchin, R. The data revolution: big data, open data, data infrastructures and
their consequences. (London: Sage, 2014) [ISBN 9781446287484].
Parker, G.G., M.W. Van Alstyne and S.P. Choudary Platform revolution:
how networked markets are transforming the economy and how to
make them work for you. (New York: W.W. Norton & Company, 2016)
[ISBN 9780393249132].
8
Ryan, J. A history of the internet and the digital future. (London: Reaktion
Books, 2013) [ISBN 9781780231129].
This book is a very readable and chronological history of the internet, with
lots of sources of data referenced. We recommend in particular Chapter 8
on ‘The Web’.
Yoo, Y. ‘Digital materiality and the emergence of an evolutionary science of
the artificial’ in Leonardi, P., B. Nardi and J. Kallinikos (eds) Materiality
and organizing: social interaction in a technological world. (Oxford: Oxford
University Press, 2013) [ISBN 9780199664054] pp.134–54.
1.6.4 Online study resources (the Online Library and the VLE)
In addition to the subject guide and the Essential reading, it is crucial that
you take advantage of the study resources that are available online for this
course, including the VLE and the Online Library.
You can access the VLE, the Online Library and your University of London
email account via the Student Portal.
You should have received your login details for the Student Portal with
your official offer, which was emailed to the address that you gave
on your application form. You have probably already logged in to the
Student Portal in order to register! As soon as you registered, you will
automatically have been granted access to the VLE, Online Library and
your fully functional University of London email account.
If you have forgotten these login details, please click on the ‘Forgotten
your password’ link on the login page.
The VLE
The VLE, which complements this subject guide, has been designed to
enhance your learning experience, providing additional support and a
sense of community. It forms an important part of your study experience
with the University of London and you should access it regularly.
The VLE provides a range of resources for EMFSS courses:
• Course materials: Subject guides and other course materials
available for download. In some courses, the content of the subject
guide is transferred into the VLE and additional resources and
activities are integrated with the text.
• Readings: Direct links, wherever possible, to essential readings in the
Online Library, including journal articles and ebooks.
• Video content: Including introductions to courses and topics within
courses, interviews, lessons and debates.
• Screencasts: Videos of PowerPoint presentations, animated podcasts
and on-screen worked examples.
• External material: Links out to carefully selected third-party
resources.
• Self-test activities: Multiple-choice, numerical and algebraic
quizzes to check your understanding.
• Collaborative activities: Work with fellow students to build a body
of knowledge.
• Discussion forums: A space where you can share your thoughts
and questions with fellow students. Many forums will be supported by
a ‘course moderator’, a subject expert employed by LSE to facilitate the
discussion and clarify difficult topics.
9
• Past examination papers: We provide up to three years of past

examinations alongside Examiners’ commentaries that provide
guidance on how to approach the questions.
• Study skills: Expert advice on getting started with your studies,
preparing for examinations and developing your digital literacy skills.
Note: Students registered for Laws courses also receive access to the
dedicated Laws VLE.
Some of these resources are available for certain courses only, but we
are expanding our provision all the time and you should check the VLE
regularly for updates.
Making use of the Online Library

The Online Library contains a huge array of journal articles and other
resources to help you read widely and extensively.
To access the majority of resources via the Online Library you will either
need to use your University of London Student Portal login details, or you
will be required to register and use an Athens login.
The easiest way to locate relevant content and journal articles in the
Online Library is to use the Summon search engine.
If you are having trouble finding an article listed in a reading list, try
removing any punctuation from the title, such as single quotation marks,
question marks and colons.
For further advice, please use the online help pages or email the Online
Library team.
Each chapter of this guide refers to electronic articles and texts which you
can reach through the VLE and Online Library, or directly on the internet.
In addition, the VLE has many videos about all the topics in this guide
and beyond. We expect you to explore the readings and videos in order to
have a better understanding of the topics. The VLE is regularly updated
with content. It is fundamental for your success in this course that you
engage with the online content. It allows you to learn the topics in greater
depth. It also helps you to find interesting content to discuss with your
teachers and fellow students. Make sure that you explore the material for
each chapter as the course moves from topic to topic. Thus, the learning is
better structured across the course. Observe though that some links may
not work through time, as the content on the internet changes a lot. If a
link is broken, you can find the content online by yourself, using the search
engines such as Google (for written material) or YouTube (for videos).
Alternatively, you can alert us to a broken link via the Student Portal.
10
Chapter 2: Infrastructures: conceptual beginnings
Chapter 2: Infrastructures: conceptual

beginnings
Chapter outline
This chapter introduces the concept of infrastructures. It discusses the
concept itself and emphasises the importance of infrastructures in the
development of forms of production in the economy and organisation
of society. We argue here that just as roads and railways and electricity
grids were core infrastructures of the 19th and 20th centuries,
digital technologies have become fundamental infrastructures in our
contemporary world. Of course, roads, railways and electricity are still
important for us, but the digital infrastructures discussed here are at the
heart of how society and the economy develop today. We argue that the
emergence of digital infrastructures has serious implications for how
well countries, industries, corporations and people make use of digital
technologies, devices and services. This chapter introduces examples of
digital infrastructures which are discussed across the course. In particular,
we introduce here five examples of digital infrastructures: the internet, the
WWW, cloud computing, mobile technologies and social media. Later in
this subject guide each of these examples are discussed in more detail in
their own chapter, using the theoretical lenses presented in the guide.
2.1 Introduction
From a historical perspective, infrastructures have long been seen
as important for economic and social development. For example,
the establishment of railways in the USA in the 19th century, or the
constructions of separate fresh water and sewage systems for European
cities in the same century were fundamental to economic and social
progress. In the 20th century it is perhaps the road system, telephones
and the electricity grid that we can identify as the outstanding and
fundamental infrastructures that drove change. In our own era, in the
third decade of the 21st century, digital technologies and the services they
provide are understood as similarly fundamental infrastructures for our
societies and our economy.
Developing and using digital infrastructures has significant implications
for all countries and their governments often shape policies to promote
their development while sone companies make the provision of digital
infrastructure their primary business. This emphasis on infrastructures
is often justified by a belief that infrastructures drive the adoption of
digital technologies that in turn provide economic and social benefits.
For example, when developing countries in sub-Saharan Africa adopt
smartphones, they thereby stimulate all sorts of economic and social
improvements. Because of the existence of a good mobile phone
infrastructure, including phone masts for wider coverage and pay-as-
you-go phone rental, new services can be offered and new businesses can
be developed. Similarly, digital entrepreneurs can start new businesses
in London, São Paulo, Shanghai, or Singapore, in part because these
cities provide excellent 4G and broadband availability. And, of course,
other cities and other countries are also eager to find similar economic
development opportunities.
11
2.1.1 Aims of the chapter

• Conceptualise infrastructures and digital infrastructures, providing
examples.
• Demonstrate how infrastructures affect production models and forms
of organising.
• Show how digital infrastructure became essential in contemporary
societies.
• Explore examples of how digital infrastructures enable new products,
services and business models.
• Explain how digital infrastructures relate to topics covered in
subsequent chapters.
2.1.2 Learning outcomes

By the end of this chapter, and having completed the Essential reading and
• describe the concept of infrastructures and digital infrastructures,
providing examples.
• explain the link between infrastructures and types of production and
organising models.
• outline the emergence of digital infrastructures and discuss its
implications for economies, corporations and society, with focus on key
digital infrastructures.
• describe examples of products, services and business models enabled
by digital infrastructures.
• explain the connections between digital infrastructure and key topics
of this course at an introductory level.

[ISBN 9781292417752]. Chapter 5, Section 5.1.
London: Yale University Press, 2009) [ISBN 9780300151244] Chapter 4
‘The generative pattern’ pp.67–100. You may find a copy of this book online
through Google Search or Google Scholar.
Similar content is available in previous and later editions of the same
book. Read the chapter in other editions which discuss the topic of
infrastructure.
Wikipedia
Infrastructures: https://en.wikipedia.org/wiki/Infrastructure
Critical infrastructures: https://en.wikipedia.org/wiki/Critical_infrastructure

Carr, N. The big switch – rewiring the world, from Edison to Google. (New York:
W.W. Norton and Company, 2008) [ISBN 9780393062281].
On the link below, Nicholas Carr presents some of the key ideas of his book,
The Big Switch: https://www.youtube.com/watch?v=Fb362AtseSA
Internet: https://en.wikipedia.org/wiki/internet
WWW: https://en.wikipedia.org/wiki/World_Wide_Web
URL: https://en.wikipedia.org/wiki/Uniform_Resource_Locator
Mobile subscriptions: http://data.worldbank.org/indicator/IT.CEL.SETS.P2?en
d=2014&start=1960&view=chart
12
Social media: https://en.wikipedia.org/wiki/Social_media

Web 2.0: https://en.wikipedia.org/wiki/Web_2.0
Smartphone: https://en.wikipedia.org/wiki/Smartphone
2.1.5 References cited

Hanseth, O. and B. Bygstad ‘Flexible generification: ICT standardization
strategies and service innovation in health care’, European Journal of
Information Systems 24 2015, pp.645–663.
Standage, T. Writing on the wall: social media – the first 2,000 years. (London:
Bloomsbury Publishing, 2013) [ISBN 9781408842065].
In the video below, Tom Standage presents key ideas of his book, Writing on
the wall: www.youtube.com/watch?v=7J0I1V5XNKI
2.1.6 Synopsis of chapter content

This chapter introduces and defines the concept of infrastructure. It
then argues that digital services have become important infrastructures
in contemporary economies. The concept of infrastructure is presented
first through an example of power suppliers and then the case of specific
digital infrastructures; the internet, the WWW, cloud computing, mobile
technologies and social media. The chapter presents examples for
facilitating the understanding of the impact of digital infrastructures on
businesses and citizens, through showing products, services and business
models which have been enabled by these infrastructures. The chapter
introduces concepts at a basic level, following a strategy of gradually
building the necessary knowledge to advance to more complex ideas
and analysis. The chapter also develops the proposition that digital
infrastructures are enablers of digital business innovation, an argument
which is developed across this subject guide.
2.2 Chapter content

2.2.1 Conceptualising infrastructures
We start by looking at the formal meaning of the word ‘infrastructure’ in
wider English usage.
The Cambridge Dictionary defines infrastructures as ‘the basic systems
and services, such as transport and power supplies, that a country or
organisation uses in order to work effectively’. Similarly, the Oxford English
Dictionary defines infrastructures as ‘the basic physical and organisational
structures and facilities (e.g. buildings, roads, power supplies) needed
for the operation of a society or enterprise’. Both definitions reflect
fundamental aspects of the concept and are relevant to the topic and
perspective of this guide.
First, they highlight that infrastructures form the base for the
operations of a given organisation, society, or country. In other words,
without particular infrastructure, those operations either will not be
possible or would be done in a less efficient or different ways. So, we can
say that infrastructures enable particular ways of doing things.
Sometimes different infrastructures compete as a means of doing things
– for example railways (the railway infrastructure) may compete with air
travel (in Europe at least).
Second, the definitions highlight that infrastructures are systems,
services, facilities, including physical and organisational
structures. In other words, infrastructures are the physical materials
used to create a base for operations, but also include the social and
organisational arrangements we agree upon to allow those physical
materials to effectively enable particular activities.
13
Third, these definitions give examples of things that are considered to

be or described as infrastructures, such as transports systems (roads),
power supplies (electricity) and buildings (airports, power stations).
Infrastructures also have the characteristic of working on a large scale
and generate benefits because of their scale and that they are shared
among multiple users. We can infer then that scale matters when
thinking about infrastructures, as does the ability to be used or shared by
different types of users. We should also see that people or businesses will in
general choose to use infrastructures which are from their perspective more
efficient, reliable, secure, cheaper, easier, or safer than alternative ones.
In sections 2.2.4 and 2.2.5 we explore these three aspects further, arguing
that information systems have become infrastructures in contemporary
societies because of their scale, the combination of material and
organisational structures and the capacity of enabling different, novel and
more efficient operations and activities.
2.2.2 Examples of infrastructures: electricity and rail networks

We live in societies that rely on infrastructures. A modern city could not
exist without common, shared infrastructures (electricity, water, sewerage,
subways etc.). You could almost say that a city is made of infrastructure.
Look around you and the community that you live in. You probably have
electricity, running water, roads and train or bus networks, to cite a few
examples. You also have communication networks, such as telegraphs,
telephones and the internet. And each of these examples of infrastructure
have their own organisational element – the phone company, the water
department and the bus company.
Let’s start by considering in a little more detail one ubiquitous example:
the infrastructure that supplies electricity.
Today most of us take for granted that electricity is always available,
although we should remember that some Worldwide students live in places
where electricity is less than 100% available or reliable. We enter our
houses, offices and schools and we expect to be able to turn on the lights
and to have (many) sockets to connect our computers and mobile devices
(and note: all those sockets will be the same in a particular country and
reflect a common standard. We can think of a plug and socket design
as being a standard interface between your device, for example, a radio
and the electricity infrastructure).
As we know, easy and ubiquitous availability of electricity is possible
because we have power stations generating the electricity we need. This
electricity is then transported from the place of production to the point
of consumption, through a physical network of cables and distributors.
Each house and building connects to this physical network, following
established standards of wiring, switches and plugs. We can also say that
electricity is a ‘commodity’, a general resource that different people will
use for different purposes – for me to play music, for you to cook food and
for a business to run machinery. Although electricity may not be exactly
cheap in most places, it is certainly affordable considering its cost is much
lower than the benefits it brings. In this way, the electricity supply network
– the electricity infrastructure – brings a huge amount of benefit to society
by allowing uncountable activities, from preserving food and warming
or cooling buildings, to extending working hours, allowing the use of
industrial machines and household devices and improving the security and
safety of roads at night. Think what happens on the few occasions when
communities used to reliable electricity supply have to face a power cut or
blackout.
14
Before we had this power supply infrastructure, the situation was not so
comfortable. Even when the technology for producing electricity had been
developed, very few companies and households had the resources to adopt
the new kind of power. Companies and households could install equipment
for producing electricity (generators) and use this locally – but supply
equipment had to be installed very close to the place of consumption. In
that period, electricity was a great invention but one that only brought
advantage for a few who could afford it, but not much for the rest.
The situation changed as governments and entrepreneurs came to
understand that larger numbers of people and businesses could benefit
from having access to an electricity supply. However, in order to make this
possible, it was necessary to change the model for generating electricity.
Instead of having a myriad of small local plants, it was necessary to have
larger, centralised units to produce electricity at a lower cost (economies
of scale). These units would not in general be local, so it was necessary
to build a network of cables to deliver electricity across large distances
and distribute it to multiple sites. There was then the problem as to how
to deliver the electricity for each house or building. Technical standards
and rules of operation needed to be agreed among industries, technical
bodies and governments, allowing the connection of individual buildings
and houses to the network, in a safe and regulated manner. And of course,
there was a need for an organisational structure too, the electricity supply
facilities (power stations, cables etc.) needed to manage on a day-to-day
basis, bills needed to be issued and new investments made to upgrade and
grow the electricity system.
Having a standard ‘kind’ of electricity (for example 230v, 50Hz in a
European country, but 110v, 60 Hz in the USA/North America) and a
standard kind of plug and socket made the social and organisational
arrangements easier too. For a worldwide overview of all the different
electricity standards we have today see the plug voltage by country. There
are still a surprising number. With agreed standards, consumers in a
country can go into a shop and easily purchase a refrigerator ready to be
plugged in and work (what today in computers is sometimes called ‘plug
and play’). Manufacturers also benefit since they only need to make a few
models to supply the whole world, even if they do need to put different
plugs on the cables. And both consumers and manufacturers benefit from
having regulations that define quality and safety standards, plugs and
sockets and rules for pricing.
The building of the power supply networks has taken a long of time –
over 140 years in the UK. The original model of individual power supply
units was gradually substituted over decades by the centralised and
shared resource model. New services were offered first in a few cities and
gradually increased places were integrated to the network of power supply.
The scale has become huge, reaching national levels and interconnecting
networks across countries when the production of electricity is done in one
country and sold to another. Importantly, as the scale has become as bigger
it has allowed a reduction in the cost of electricity, benefiting all in the
same network.
So here we are today in a world when most of us, but not all of us, take it
for granted that our buildings have plenty of electricity available on tap,
thanks to an established electrical infrastructure.
However, we need to understand that the emergence of infrastructures
is not always smooth. When the power supply was provided locally, the
technical arrangements were done locally too, thus it was difficult to
15
migrate from lots of different local arrangements to a large scale and

uniform supply network. It was necessary to discuss and agree common
standards. Doing this it is certain that conflicts will emerge, about what
sort of standards and rules should be followed and which choices should
be made and on what criteria. This leads us to another fundamental aspect
of infrastructures. Most infrastructures are not planned for in advance on
the basis of a clean sheet of paper and a fresh start. Rather, they emerge
out of pre-existing arrangements and resources (what we call the installed
base). They emerge from pre-existing systems and services, trial and error,
argument and false starts. Still, the impetus to develop infrastructures is
strong because the logic of scale, of standards and of sharing drives in
forward as well as the innovators who find new uses for the infrastructure.
We can say that infrastructures emerge from technological advances and
entrepreneurial innovators, both of which help them spread across societies,
but this is seldom a simple case of ‘onward and upward’. One example of
the electrical infrastructure evolving that we see today in many countries is
the further development of electrical infrastructures (in combination with
digital infrastructures) to support electrical vehicles (EV).
This reliance on the installed base has one big and important implication.
Any current or emerging new infrastructure probably has to be built on and
make use of older resources and systems that went before. And doing this
does not necessarily lead to infrastructures that operate or are structured in
the most efficient way. In other words, the development of infrastructures
has a path dependency with previous modes of operation and resources
limiting the way infrastructures evolve through time.
Activity 2.1: How train networks and water supply have become national
infrastructures
You have learnt in this section how power supply networks became common and shared
infrastructures. In this activity, you do your own research to understand how either rail
networks or water supply have evolved as infrastructures. You can research the process
in your own country, or you can investigate the case of national rail network in another
country – the United Kingdom or the USA for example.
You should first establish, based on the ideas discussed above, to what extent the
railways system or water supply is worthy to be called a national infrastructure – perhaps
more in some countries than in others. Then you should find two or three examples of
how disparate and ‘local’ rail or water supplies have been transformed and integrated to
become part of a larger whole with common characteristics and economies of scale.
The objective of this activity is for you to better understand how technology innovation
will sometimes evolve to create an important infrastructure. In the United Kingdom,
the rail network was originally fragmented. Each part of the network followed different
standards, including the width of tracks and the timetables. Moving across the
network was very difficult, engines and trucks could not pass around the country and
manufacturers had to serve the multiple needs of different railways asking for different
products. It took a lot of time and effort to make the fragmented rail network something
worthy to be called a national infrastructure.
This activity also helps you to develop your research and self-study skills, which are
fundamental for preparing you for continuing learning.
2.2.3 Infrastructures, production and organisation

In section 2.2.1, we emphasised that infrastructures enable particular
activities and operations and forms of organising, offering services that
are more efficient than alternative (e.g. local, small scale) ones. Indeed,
a particular ‘solution’ emerges as a shared and scaled-up infrastructure
16
because of its capacity of bringing value to organisations and consumers

– be it as efficiency, reliability, accessibility, security etc. The best way of
understanding this process is through examples.
We start with the power supply networks, discussed in the previous
section. The use of electricity has changed the world in many ways. As
discussed before, we can today be productive during the whole day, not
depending on the sunlight, or on other less efficient technologies such as
the oil lamp. We can use electric machines to produce products in a more
efficient and clean way than the machines driven by steam. Indeed, a
great variety of machines and devices have become possible only because
electricity is available. Have a quick look at your home. You are going to
find many devices which depend on electricity. What are the benefits for
your life in terms of productivity, security, leisure and health?
Similarly, rail networks have brought substantial gains in productivity
and wellbeing for societies. Before the development of the railways,
movements between places were slow, depending often on the power of
animals. The train offered a clear advantage compared with horses in
terms of productivity and capacity. However, this was only really possible
when trains are used across the country and across countries. It is not
a coincidence that many of the first rail networks were built to connect
industries and farms to ports, in order to transport goods far away from
their locale of production. Think about England in the 19th century,
building the first rail networks to connect its industries to key ports,
making it possible to export products to many other countries and to
develop larger industrial enterprises.
Similarly, many countries today invest in their port infrastructures to
support industrial development and the maritime network is an important
infrastructure that allows countries to reach other markets. This maritime
infrastructure depends on ports, shipping routes, all sorts of ships to
transport different goods and a particular form of organising business
to allow these operations, including national and international laws
and payment systems. We can see that infrastructures are also often
connected to each other and together may build an even more effective
infrastructure. In this case, we can think, for instance, about transport
infrastructure, which puts together all means of transport of a country,
from cars and buses to trains, trucks, ships and airplanes.
Finally, let’s consider communication networks. You may have read
about the old times when people would communicate by letters dropped
in the postbox. Not so long ago people could not send a telegram, or
make a phone call, or if they did it would cost serious money. But with
the advancement of communication technology, societies have seen the
benefits of investing in building communication infrastructures. The bigger
the number of people using communication networks, the bigger the
benefits for each of the users of these networks (what is called a positive
network effects). If only a few people have telephones, the cost of
maintaining the service is high for users and their benefit limited. However,
when most people can afford to have a telephone, the cost for individuals,
households and companies is reduced (economies of scale) but also the
benefits they see are increased because they can contact more people. This
idea of a positive network effect says that the more people using a
network or infrastructure, the more value they each derive from it.
In order to better understand the impact of infrastructures in the
production and organisational models, we can compare two scenarios of
places with different infrastructures.
17
Scenario 1:
John produces vegetables in a small village with poor power supply,
transportation and communication infrastructures. Like any small farmer,
he struggles with the weather to get good quality products which can
reach the market. As he does not have a good power supply, he cannot
afford to have refrigerated rooms for the products, implying he needs
to sell the production very quickly. The quality of his vegetables could
be better with a little bit of mechanisation, but so far this has not been
possible as he cannot rely in the supply of electricity. However, John
faces other obstacles. Living in a small village means he cannot rely on
the transport network to get his product to the market in the closest big
city. Sometimes the roads are working, sometimes not. It depends on
how much the rain season has destroyed the roads. John would like very
much to have further markets, but this is not possible. In addition to not
have means for transporting the goods, he also does not have means to
communicate efficiently with other markets or wholesalers. The telephone
network is very poor in his village and it is expensive as well. He even
cannot afford to regularly communicate with the retailers in the big city to
negotiate the price of his goods. He depends mainly on a single wholesaler
who visits the village once per week – when the weather helps! – to buy
his products. And John accepts the price offered by the wholesaler as he
does not have an alternative.
Scenario 2:
Anne produces vegetables in a small village which is well connect to power
supply, transportation and communication infrastructures. As electricity is
accessible and affordable, Anne has invested in mechanical irrigation and
refrigerated rooms to keep her products fresh for longer periods. Through
the internet, Anne gets all information she needs to better cultivate her
vegetables and to know exactly what the markets are willing to buy. She
has segmented her production to different markets. Most of her production
goes to the nearest big city. She is informed daily about the prices in the
market and can send fresh products regularly, using the efficient transport
networks. The intermediary makes a fair amount of money, but he is not
able to manipulate prices, as Anne has information about the prices and
demands of each of her products. The other part of Anne’s production is
exported. Knowing about the demands in other countries, she has joined
a network of small farms which collaborates to get the fresh products
from the farm and transport them by plane. In a matter of few days, fresh
products can leave Ann’s farm and reach households and restaurants in
other countries. Ann monitors the prices and the markets daily and she
can quickly redirect her production either to the nearest city or to other
countries, getting the benefit of the best price.
The examples show in a very small scale how and why infrastructures
matter. The more countries invest in energy, transportation and
communication infrastructures, the more they have advantages in creating
more competitive and efficient production models. It is not only about the
use of technology, such as in our example the use of refrigerated rooms
and better transportation systems. It is also about the quality and timing of
information to make decisions. The forms of organising production, with
a fairer distribution of wealth among entrepreneurs in the value chain
and better access to different markets, can improve both competitiveness
and fairness in society – as for example giving small rural farmers a better
opportunity to sell their products and make profits to reinvest in their
enterprises.
18
2.2.4 The emergence of digital infrastructures

From the general understanding of infrastructures introduced above
using examples of electricity, roads etc., we now turn to the emergence
and relevance of digital infrastructures and their impact in societies and
organisations. We start with an infrastructure that everybody knows to
some extent: the internet.
The history and technical aspects of the internet will be discussed in detail
later in this guide in Chapters 3, 4 and 5, but we can anticipate here some
important aspects.
The internet became a commercial technology available to all to
potentially use in 1993. Before this, only some groups of scientists
and researchers had access to the internet. Before long, companies,
governments, civil society organisations and individuals realised the
enormous potential of the internet and just how useful it could be for all
sorts of tasks and activities.
In the beginning, the internet was mainly used for the exchange of files
of data and then emails – a huge gain of productivity in relation to other
communication channels such as letters and memos. It also soon became
used for the publication information – principally via the early versions of
the WWW. At that stage, the internet was an interesting technology and
spurred interesting applications, but it was not yet perhaps an important
national or global infrastructure. The situation changed over a short
space of time, as the internet (and in-company intranets) became a major
channel of communication (email, online chat), information sharing (the
web) and most importantly, a place to do business (e-commerce).
The more the internet gained speed and capacity (broadband) the more
it could accommodate interactions between people and people or people
and businesses and thus allowed electronic commerce for buying and
selling products and services online, It could and support communication
(such as voice over IP (VOIP) and video-conference calls) and diffusion of
content (websites, blogs, videos etc.). As these uses expanded, reinforcing
each other (more of one led to more of the others), the more the internet
became an important infrastructure for economies and societies across
the world. All those different uses and more and more people worldwide,
started to share the communications capacity of the internet and use it
for their own diverse purposes.
An easy way of understanding this is to think about not using the internet
for a week – which means not using most of the apps on your phone. How
many of your activities are going to suffer disruption? How many of your
activities are going to lose efficiency and/or cost more? Do you know the
alternative ways of doing the things you do on the internet if the internet
is not available? Chances are that some activities would just not be
possible at all if you could not use the internet for a whole week.
The internet is the most fundamental example of a digital infrastructure
which has changed the way we do things. However, the internet became
commercially viable initially only because it made use of (or we might say
was ‘built on top of’) the then existing communication infrastructure of
the telephone network. Indeed, as we noted above, most infrastructures
emerge partly from design and planning and partly from borrowing/
building on existing systems or services and putting them to new purposes.
Even the railway companies in Great Britain in the 19th century bought up
canals (the infrastructure of the previous century), filled them in and laid
railway tracks on them.
19
As outlined in Chapter 3 below, the internet was initially designed to

transfer digital data through then existing communication networks,
what we call legacy systems or the installed base – the systems we
inherit. By linking together diverse networks and resources the internet
became a bigger network with more individual connections and remember
more connections, more people creates more value for all. Where digital
infrastructures evolve from previous legacy systems, this sometimes
creates a lock-in situation. This occurs if the whole infrastructure
cannot change because some of its existing parts are dependent on
legacy elements and the infrastructure needs to accommodate older or
slower technologies. Change in a digital infrastructure is expected as
infrastructure take up new technologies and improve their performance,
but it can be difficult to achieve because of this kind of path-
dependency – defined by Wikipedia as when ‘the set of decisions one
faces for any given circumstance [is] limited by the decisions [ …] made in
the past, even though past circumstances may no longer be relevant’.
On the basis of this introduction to digital infrastructures and the concepts
introduced in the earlier part of this chapter, we suggest a basic set of
six characteristics that we will want to consider in any and every kind of
digital infrastructure we study, as described in Figure 2.1 below. We will
return again and again to these ideas throughout this course.
• Shared: a digital infrastructure is shared by the members of a community who use

it for their own purposes.
• Heterogeneous: a digital infrastructure consists of different elements:
technology, users, organisations.
• Installed base: a digital infrastructure will build on, expand and integrate
previous systems and services.
• Evolving: a digital infrastructure is not designed or created in a single act but
evolves continually as it expands in scale and is adapted for new purposes.
• Open: a digital infrastructure is available for new users and new uses.
• Standardised: a digital infrastructure needs agreed common standards to ensure
its functional integrity and allow it to scale up.
Figure 2.1: Six core concepts for identifying a digital infrastructure (Adapted
from Hanseth and Bygstad, 2015)
2.2.5 Digital infrastructure and business

Considering a supermarket chain like Walmart can help us to understand
the impact of a digital infrastructure on businesses. Put very simply, a
Walmart supermarket needs to do three main activities: buy products,
sell products and decide on future investments. In order to buy products,
Walmart relies on supply chain management software, which allows the
firm to know immediately when it is necessary to replace the stock, to
procure better and cheap products and to electronically manage the logistics
to get the products on time in its stores. In order to sell products, Walmart
relies on digital marketing to promote its products and collects relevant
information about customers (such as using a loyalty card), on checkouts
which read bar codes and in internet interfaces to provide better customer
services (integrated into customer relationship management software). In
order to make decisions on future investments, Walmart relies on business
intelligence which applies data analytics upon big databases (Big Data,
discussed later in this subject guide) to forecast which investments will
20
generate better returns. Based on this data, the supermarket chain may
make decisions on where to open new stores, which products and services
will sell better in particular stores and how to perform the logistic and the
operations of the day considering the needs of each store.
Focusing on the software aspect, we can see that a supermarket chain such
as Walmartm relies on different pieces of technology, such as supply chain
management, customer relationship management, the internet, databases
and data analytics. With a little bit of research, we could find that indeed
the whole Walmart business is integrated into an enterprise resource
planning software, which integrates all these elements with others,
which manage, for instance, the financial and human resources aspects
of the company. We could also find that a variety of interfaces is used for
digital marketing, including Google, Facebook and owned channels
(e.g. Walmart’s or Amazon’s own pages) and that many different digital
tools are used to explore sentiments and customers’ disposition to spend
(individually or in clusters).
In other words, we can observe that Walmart depends on a complex and
extensive digital infrastructure to make its business possible and profitable
and that this infrastructure has many elements and many layers that work
together to deliver more value to the supermarket’s shareholders. Some of
these are internal and some are external to the organisation.
Walmart can only make this extensive use of digital systems because there
is a bigger network that connects the firm to other businesses, which also
use similar pieces of software and decide to communicate in the same way.
Walmart can adopt digital marketing because its potential customers are
online, using Google and Facebook and researching prices on different
websites. Walmart can use the internet to improve customer service
because consumers go online to put forward their complaints and expect
to receive feedback using the same channel. So Walmart creates for itself
an infrastructure that is embedded in a larger digital infrastructure (the
internet, Google, Facebook, the world wide web etc.), which is a legitimate
business base for any number of other companies. Indeed, companies that
do not adopt such digital infrastructures increase their chances of going
out of business as the market moves towards digital environments.
Activity 2.2: Key ideas behind the concept of digital infrastructure

On the basis of the sections 2.2.1, 2.2.2 and 2.2.4 and your further reading, prepare
an initial version of a document entitled: Digital infrastructure: key underlying
concepts. As you read more of this subject guide and other references, update and
develop this document.
2.2.6 Innovation and digital infrastructure

Innovation is a word that is very familiar today. It is usually used in a
positive way by businesses, entrepreneurs, economists, governments and
scientists and it is even in the title of the degree that this course is a part
of. The general sense in all these uses of the word is that innovation is
good, it’s what we need to power our economies and to make life better
for people. In this sense innovation means new things (products and
service), new ways and new places for doing things (processes and places)
and new ways for people to think (perceiving) and act (behaviours). These
new or changed aspects of our world will then, we expect, improve our
lives by creating economic value, human wellbeing and prosperity.
21
The more globalised and competitive the world is and the shorter product
life cycles are, the more relevance being able to innovate is seen to have for
businesses and economics. Keeping making profits, holding market share
and sustaining the relative capacity for generating wealth in an environment
of globalisation and climate change demands that economies keep up the
process of innovation. This is true for national or regional economies and
for individual firms looking for a sustainable/viable future. We expect
other positive benefits from innovation too, independently of them keeping
companies profitable or countries economically healthy. These may include
things such as environmental benefits (innovation in green technologies),
social wellbeing (innovation in educational technologies), health (innovation
in medicines, diagnostics and treatments), happiness and safety (innovation
is social and cultural norms).
There is a view that sees innovation – change in how things are done
and are understood – as the central driver of our economies; what the
economist Joseph Schumpeter called the ‘gales of creative destruction’. Out
with the old, on to the new. His idea was that pursuing new or different
things (products) or new different ways to do things (processes) created
opportunities for more and different kinds of economic activity. As new
ideas gain traction in the economy, inefficiencies or vested interests
are swept away in the gale, leading to a different more productive and
perhaps socially more agreeable future. Today we might see these gales in
areas such as the move to a new greener and low carbon economy, or the
challenge that Fintech companies pose to traditional banks.
Tidd and Bessant (2013, p.19) present many definitions of what
‘innovation’ means. One that is very short and precise is: ‘Innovation is the
successful exploitation of new ideas’ (as defined by the Innovation Unit, UK
Department of Trade and Industry, in 2004). Innovation using this approach
is not principally about having new ideas or even inventing something;
rather, innovation is the process of growing new ideas into practical use
(Tidd and Bessant, 2013). Following this line of thought, the process of
innovation starts in searching for an interesting pool of ideas, in selecting
from this idea which seem more promising and then trying to put these
ideas to some practical and value-adding use. This is, in a general way,
what biotech industries and start-ups do or what venture capital funds seek
to achieve. This view of innovation certainly fits with the account of the
development of the internet as set out in Chapter 3 of this subject guide.
We might also include a final stage of innovation as capturing value from an
innovation – remembering that there is not always a simple and direct link
between who leads the innovation, who benefits from it and who gets rich!
Tidd and Bessant (2013) go on to suggest that innovation can happen
in four key areas, rather similar to the ideas introduced in the opening
paragraph of this section. The most obvious one they suggest is innovation
in products and services themselves. The second is innovation in processes
(i.e. changes in the way products and services are created and delivered).
The third is innovation in position, when the innovation changes the
context in which products and services are introduced. And the last area
is innovation in paradigms, when the changes affect the mental models
(including business models) which define the way organisations or social
and economic systems works. Each of these four areas of innovation are
relevant for our goal of better understanding the dynamics of digital
infrastructures and their impact in organisations, markets and society. Of
course, any given innovation may have consequences in more than one of
these four areas.
22
Activity 2.3: Key ideas behind the concept of digital infrastructure

1. Think about the things you do day-to-day on your mobile phone. Identify different
apps and services that you use that clearly represent each of the four kinds of
innovation that Tidd and Bessant write about. As noted above, a single app may
represent a mixture of kinds of innovation. You could possibly argue for any or all of
these categories, but which one does it best represent for you?
You can also ask the opposite question. Is there anything on your mobile phone
that you use that is not an innovation, that represents a continuity in how the world
works, what we do and how we think?
2. Think about four large digital service providers (digital business platforms) that you
use of are familiar with (e.g. Amazon, PayPal, Facebook TikTok, TripAdvisor, Zoom,
Slack, WeChat, Alibaba, Vkontakte, Odnoklassniki, Office 365 etc.). You are not
confined by this list of examples – it is better to make your own choices of services
you use and major players in your local environment.
In each of your four cases, identify what you think is the core type of innovation
these service providers have managed to harness to become successful. In the
spirit of Schumpeter and the ‘gales of creative destruction’, suggest what is lost or
blown away in each case. For example, we might say that PayPal and other digital
payment services are a process innovation that changes the way we do a very basic
activity – paying for goods. A service such as TripAdvisor might be seen as a paradigm
innovation – changing how we (consumers) think about travel and how the managers
of hotels and restaurants understand success in their jobs as managers.
The ideas on innovation introduced here are used across this subject
guide. We explore how particular technologies can be understood as
innovation both in the development of digital infrastructures themselves
(e.g. WWW and internet) and in what they come to be used for – a topic
discussed in depth in Chapters 11, 12 and 13. We also discuss innovation
from the perspective of new digital products and services (Chapter 14) and
innovation (doing new things, finding new ways) from the perspective of
digital business platforms and new business models (Chapters 17 and 18).
In Chapter 19 we discuss some aspects of how we perhaps can innovate
the way we live in smart cities.
2.2.7 Examples of digital infrastructures

This subject guide discusses in depth five important and related digital
infrastructures: the internet, the WWW, cloud computing, mobile
technologies and social media. Each has one or more chapters devoted to
it in this subject guide. We say they are interrelated for two reasons. First
as you can probably know, the internet is the infrastructure upon which
the others rely. The WWW is a distinct information infrastructure and it is
built on and uses, the functionality of the internet. Social media such as
Twitter and Facebook are worthy of being considered as an infrastructure
in their own right too (social infrastructures), even as they depend on
and use the internet for their communications needs and the WWW as
their interface to users. Thus, we might say, simplifying somewhat, that
the WWW is an information infrastructure, social media is a social
infrastructure, while the internet is a communications infrastructure –
but all three are examples of the broader category of digital infrastructure.
The internet
In 2021, the internet was used by around 63% of the worldwide
population, i.e. 4.9 billion people had access to the internet (International
Telecommunications Union, ITU). Every day more people get access thus
23
this number is growing. However, this number also shows that 2.9 billion
people still do not have access to the internet, with all the negative
consequences of being on the wrong side of the digital divide.
You probably have quite a good understanding of the internet from
the perspective of a user. You probably also understand that the other
infrastructures that we introduce here (WWW, Cloud, Mobile and social
media) are all themselves developments that use the internet. One way
to express this is to say that they all sit ‘on top’ of the internet. You could
also say that they are created out of the internet and that the internet is
a key part of their installed base. We will come back to this idea of one
infrastructure ‘sitting on top’ of another more formally in Chapter 6 and
explore the idea of an ‘installed base’ in Chapter 7, but for the moment we
ask you to take a common sense understanding of what we mean.
The internet itself is a network of connected computers and computer
networks, which extends across the world – a network of networks. The
internet is a complicated kind of thing made up of:
• physical hardware and software (computers, cables, satellites,
programmes, devices etc.)
• a common set of rules or standards that it operates by, in particular the
protocol for the exchange of data between the computers (TCP/IP)
• a complex set of organisations and institutions that govern and
manage the service and help it to develop.
What makes the internet something special and distinguishes it from other
digital communications services, is the way it connects computers using
an open standard (TCP/IP) allowing any connected computer can access
any other. Closed computer networks may allow communication within
defined boundaries (e.g. for a government, a particular firm, a sector of
industry). But the internet is open: it offers unlimited access, so long as
you are in a country which does not block content through firewalls and
filters. Anyone can add new computers to the network, as long as the
standard communication protocol is used. And the internet itself has no
concern with the content of the messages it transmits.
However, what really makes the internet unique and fundamental is the
willingness of people, organisations, institutions, governments, charities,
non-governmental organisations and any sort of organised group to pass
messages and digital content through the network, to look for content
across the network and to communicate through the network. The internet
has become the world’s preeminent digital infrastructure precisely because
all these social actors have decided to publish and find content and to
communicate through the network. And the more computers and people
there are in the network, the more dominant the internet becomes.
Today, the internet is the most powerful digital infrastructure, for its reach,
relevance and capacity of wealth creation. It is the infrastructure which
allows the exchange of emails, the online commerce of goods and services,
the interaction with companies and governments, the access to news
and knowledge of all sorts, social interaction with friends and strangers,
among numerous other activities. Think about your life today: how many
things you do on the internet? What would happen if the internet were not
available for one month?
In Chapters 3, 4 and 5, we discuss the history of the internet and the
relevance of internet protocols in detail.
24
The world wide web (WWW)

The world wide web (WWW or just web) was created by the English computer
scientist Tim-Berners Lee in 1989. It is a system for identifying documents
and other resource on an internet computer using URLs (Uniform Resource
Locators). A URL provides the address of any object in the web – such as a web
pages, photos, graphic images, forms and videos. As a metaphor, one could think
of the web as an information space in which pieces of content can be located
because they have a unique address. The web embodies the hyperlink concept
to navigate from one web page or web resource to another as a link embedded
in the information you access. You can be sure that the link https://www.bbc.
co.uk/news is going to go to the BBC news page on the internet. This simple
mechanism of hyperlinks facilitates the lives of all of us, allowing quick and
direct navigation from one kind of content to another.
We can retrieve and read/listen/look at this content, or respond to it in
a published comment, an edit of a file, or filling in a form, as long as the
presentation of all content uses standardised languages and conventions which
other computers can identify, such as HTML (Hyper Text Markup Language),
XML or JavaScript used on web pages, or the pdf format for documents. We
might imagine this as a big library with millions of books and billions of photos
and videos, but without a searchable catalogue. Would this library be useful?
Probably not. We need to have ways to find the books, photos and videos when
we need them to make the library useful. The web would thus have less value for
us all if it was not also possible to search and find specific content easily.
The solution that has emerged is found in search engines such as Google, Yahoo
and Bing – a key part of the WWW. Hyperlinks are central to the way these
search engines work. The search engines use the page content (words and
phrases) but also hyperlinks – those embedded there and those that refer there
– to define the relevance and importance of a particular page. In this way they
decide if it should ‘rank’ high and be shown to you in the first page of a search or
rank lower and be on a later page.
Finally, we should understand that the world wide web is often conflated with
the internet, as if they are one and the same. The confusion comes because these
two digital infrastructures work so well together, like two pieces of a bigger
mechanism. However, we should understand that they are indeed different with
different histories and different purposes.
In Chapter 10, we discuss in detail the web and web engineering related topics.
Cloud computing
The idea of cloud computing is simple and logical if you understand
infrastructures. It is based in the concept of computing as a utility service, just
like electricity or water. When you buy or rent a house in most urbanised places
in the world, you take for granted that there will be an electricity provider to
supply the service to your house. You do not expect to be forced to generate your
own energy. Electricity has become a utility.
Let’s think now about how this idea of utility can be used for computing. The
key concept is the same: the final user does not want to be concerned to manage
their computing needs. Most parts of the services can be provided by suppliers
and delivered by the internet. These suppliers and their customers can reap
economies of scale.
A simple example: let’s say that five or more years ago Ann, a computer user,
wanted to use a particular software package to create documents (e.g. Microsoft
Word). In order to do this, Ann bought the software on a CD and a license to
use it and she installed the software on her computer. She needed to ensure
25
the software was updated regularly, going online to download the new
versions. Ann also knew that substantial changes in the product would
not be updated for free, new versions of the product would need to be
purchased again. Ann also wanted to be sure that her computer had a
backup of her documents. In order to do this, she had to buy a physical
backup disk for the computer and she needed to do this backup regularly
to avoid data loss.
In this example, the user needed to manage the updating of the software,
to accept that new versions of the software are not going to be for free, to
invest in a backup device and to take time regularly to do the backups.
Today we have a different way of achieving the same outcome with less
work. Now our user accesses the software package over the internet.
There is no need to load any software on the PC so no need for a CD.
The software is accessed from a browser over the internet, or via a small
app. Any updates to the software are immediately available as soon as
the update is done. The service provider can also do backups of the files
produced on its servers, in real time. Thus, the user does not need to
load software, keep it up-to-date or do the physical backup to be sure the
content is protected. The service provider may well ask a monthly fee for
these services, but they may even offer it at no cost.
A good example of such a service that you may be very familiar with is the
‘free’ email offered by Google (Gmail) and the ‘free’ word-processing and
file storage Google offers as Google Docs. Microsoft offers such services as
Microsoft365. These are both services based on cloud computing – that is
providing services, software and file storage via (on top of) the internet.
Cloud computing can be defined simply as the provision of computing
services through online networks. These services may be data storage,
software or just a platform on which to run your own programs. By
sharing resources among a large user base, cloud computing providers
promise a more professional and efficient computing service, which may
also be cheaper. Indeed, most people and organisations adopt cloud
computing because it is a cheaper and easier solution to their needs.
The concept of cloud computing is developed in more detail in Chapter 11
of this guide.
Activity 2.4: The time for cloud computing to spread

Go online and research the reasons why cloud computing has become more pervasive in
the last decade. What are the key factors supporting the spread of cloud services?
Look for examples of cloud computing services available for individuals and companies.
Make a list of cloud computing services you already use or ones that you might easily
take up.
Mobile technology
Mobile technology does not need much introduction – it represents the
computer in your pocket, on your wrist or in your bag and perhaps also
in your car. Once upon a time, computers were only available as desktop
objects, in the same way as telephones, which depended on cables to
transmit information of any kind. Gradually this situation has changed,
for example with the development of light and portable laptops with
good batteries, mobile phones with data connections and, later, tablets
and a diversity of wearable computers (e.g. the Apple watch). Wireless
connections to allow such devices to connect to the internet (WiFi,
26
Bluetooth etc.) Today mobile technologies are everywhere and the trend is
for their presence to become increasingly more pervasive.
An important step in this transition has been the development of
smartphones. There is indeed a huge change between the first generations
of mobile phones and the phone of today. In allowing access to the
internet in a very easy way, through touch screens, smartphone technology
has allowed the creation of an infinity of apps, which reinforce the value
of our smartphones. It is a virtuous cycle in which the hardware enables
the development of apps and related services and because of the diffusion
of apps and services there is a strong demand for better hardware.
The diffusion of mobile technology has brought other unexpected
benefits. If we go back in time, to the 1970s, 1980s and even 1990s, a
main concern for the economic development of poorer countries was
how to provide proper communication through telephone landlines. At
that time, people thought about the enormous amount of investment
needed to get the infrastructure prepared for widespread use of landline
telephones. When mobile technology was first developed, it was not clear
that such a revolution would come to the developing world very soon.
However, history shows that with mobile technology, even the benefits of
better communication have been grasped and once people experience the
benefit, they convince others and that demand develops strongly.
In 2021, the number of mobile subscriptions were 8.6 billion, more than
the world population (International Telecommunications Union, ITU).
Many people have more than one mobile, but this number shows that the
pervasiveness of mobile technology is extremely high. In similar research,
in 2014, ITU found that 95.5% of the world population is within the reach
of a cell tower to allow the use of a mobile phone. This is an amazing
achievement which has overcome all the obstacles of the previous landline
telephone technology. Indeed, the number of residences with fixed
landline is going down since its peak in 2006 (1,261 billion), reaching 855
million in 2020 (ITU). Many individuals and families do not need fixed
lines anymore, demonstrating the relevance of mobile technology as the
key communication channel and access to the internet.
Mobile devices are empowered by broadband services. Around the globe,
the number of active mobile internet users is 4.32 billion and the mobile
internet traffic share of total global online traffic is 56.89% (Statista). The
mobile broadband internet subscription rate is 83% in 2021, with a huge
gap between developed and developing countries (Statista). For instance,
in Europe and the Americas, the subscriptions represent 105% and 103%,
respectively, meanwhile in Africa the rate is of 41% only.
The next step in this process could be the development of ‘wearables’,
such as Google glasses and smart watches. The idea behind these devices
is better interaction with our physical world, either because we can have
more information available at any time, or because these devices can help
us to collect information, we would not be able to know otherwise. For
instance, how many steps have you walked today? Is this enough or do
you need to walk some more?
In Chapter 12, we return to the topic of mobile technology and discuss not
only the devices, but how they support innovation through new apps and
services.
Social media
Social media is a term used to classify a large diversity of digital services
which allow people and organisations to easily communicate and exchange
27
contents, including texts, photos, images and videos, in addition to voice.

Among best-known social media services is Facebook, which, in 2021,
had 2.9 billion users. This is a massive number and it appears today that
Facebook may become even more pervasive in the future, despite being
blocked in China by the government. Naturally, similar services to Facebook
exist in China, such as WeChat (also known as Weixin), which had around
1.2 billion users in 2021 and is mainly used by the Chinese population.
Humans have always been social beings, immersed in social networks
(families, villages, workplaces etc.), so social networking is not new. What
is new is the capacity for doing this through digital means (read more
in Standage, 2013). In the first generation of internet tools, this sort of
interaction through digital services was not easy. The first internet services
allowed for publishing content in files (similar to paper), with some level of
interaction available through email and web chat type services. The content
online could generally only be changed by the publishers, but the other
people would not interact with the content (e.g. comment on it, or change
it).
Then around the turn of the millennium came what are called Web 2.0
technologies. Web 2.0 allowed ordinary people like us to go online and
publish our own content, share the content we find interesting and comment
on the content of others, individuals and organisations. It was this change
in software interfaces which allowed the creation and spread of what we
call now social media. Before, there were groups exchanging information
through emails, perhaps organised around specific topics in discussion
lists. People would sign up to a discussion list and receive the emails of a
particular conversation. Discussion lists are indeed one of the first digital
social media. However, the interface was not attractive or easy to use. Today
services such as Facebook, WhatsApp, Twitter, Instagram, TikTik, WeChat,
Tumblr, Pinterest, Zoom and Skype, make this kind of interaction easy and
pervasive – and they all make use of the internet infrastructure for data
communication. Billions of people enjoy their social media!
Focusing on the digital aspect of social media is important, but we should
not forget the wider subsequent changes in the way we – social media users
– interact. Social media tools have provided a new form of communication,
but it is our decision, our choice, to use social media that creates social
networks. Every time that we connect to a new person, send a message
through social media, publish content, or read the content of others, we
are acting to create the social media infrastructure. In other words, these
social media services are only as important as our effort to make interesting
contributions. Without the contribution of users, social media would be
nothing.
We should also observe that the development of mobile technology
infrastructures has speed up the diffusion of social media and vice-
versa, especially after the launch of smartphones. The availability of
social media at any place through smartphones have increased the
use of social media and this has also made the mobile device a more
interesting communication channel. This process is also connected to
the improvements made on the internet as a whole and particularly
to the offering of wireless broadband services through mobile phone
infrastructures. This shows that the various digital infrastructures are
supporting each other’s growth.
In Chapter 13, we are going to discuss in more depth social media, with
more details on tools. We are also going to discuss some implications of
social media (also known as social technologies) for organisations and
individuals.
28
Activity 2.5: The relevance of social media

Most of us use some social media such as Facebook, Instagram, WeChat, WhatsApp or
Twitter. We may also use social media functions in other services – such as TripAdvisor or
Booking.com when we share comments and give ratings. But are you aware of the whole
variety of social media tools?
To answer this question, go online and find key social media tools by number of users.
The Statista portal may help you to find this information. Use only the free information
(you do not need to pay for more sophisticate services):
• Statista, Global social networks ranked by number of users.
Now that you know the most used social media tools, try to visit each of their websites
or download their applications to your mobile to understand what these tools enable and
why millions of people are using them. Think about the benefits of using the tools as well
as the risk they may pose
2.2.8 New products, services and business models based on digital

infrastructures
Reading this far in the chapter you have probably realised how these
independent pieces of technology interact with each or combine to enable
something more powerful to happen? For instance, we would not be able
to have cloud computing and social media if we did not have the internet
to support these services. In similar fashion, the internet would not be
such an interesting or popular place if we could not interact through social
media and easily find the content we are interested in through the URLs
– which is possible only because we have the WWW technology including
search engines.
In a nutshell, we can say that these technologies support each other and
together they enable the emergence of something new and powerful. The
emergence of powerful digital infrastructures is possible because of each
of these technologies works as a layer, giving services to the next level. At
the same time, each of these layers nurtures the development of the others.
The emergence of apps for mobile phones, for instance, have made mobile
technology more necessary and have made the internet more important too.
The entanglement of these layers brings a level of complexity to the
whole digital phenomenon that we may find hard to grasp. In Chapter 6
we discuss a model which may help to improve our understanding of the
whole phenomenon of emerging digital infrastructures.
In this chapter we have argued that digital infrastructures are fundamental
enablers of all kinds of business in contemporary economies. We have
also shown that many different layers of technology can work together in
digital infrastructures to allow both better efficiency and more flexibility.
What is more interesting perhaps is the way they enable organisations to
create new products and services, reach new markets and in general and
to explore new business models – they drive innovation. This is one of the
key topics we are going to return to in more detail in subsequent chapters.
However, we need to have a first view of this topic in order to improve
your understanding of future chapters. A few examples can help to clarify
some concepts.
The examples below are of products, services and business models which
are enabled by digital infrastructures. They are not infrastructures
themselves, but they demonstrate our argument for the relevance of
digital infrastructures for opening possibilities for new products and new
businesses.
29
Apple iPod is a good example to start with. The iPod when it first arrived
on the market in the early 2000s was a piece of new technology, with a
strong value proposition: it allowed people to listen to their preferred
songs in a MP3 format. The first change then is in the product and the
way the product reaches the consumer. Instead of buying music CDs, to
be played in a CD player, the consumer would buy an electronic file to
be read by the iPod. The product changed from a material copy – the
CD – to an electronic copy. However, there is another difference: the iPod
allowed consumers to buy their songs through the iTunes service and only
the songs the consumer would like to buy. This was a huge innovation.
The traditional model for selling music used by record companies was
to bundle products together. For instance, you could by an album with
a collection of songs. The selection and bundling would be done by the
record company. The consumer could either buy all or not buy anything.
With iTunes, the consumer is empowered to choose exactly the songs they
like and they pay only for these songs.
The service has also changed: iTunes will keep a record of the songs you
have bought and it will restore the copies in case of loss. This would not
be possible when you bought CDs. A lost CD is a lost CD. This does not
happen with iTunes, Apple will keep track of the songs you have bought
in a cloud service. We can see here that Apple has created a new business
model. Traditional electronic companies would sell electronic devices
only. The recording companies would be responsible for making the CDs
available (or before, the vinyl discs). But Apple entered the space of record
companies and electronic gadget companies, unbundling the albums
and making individual songs directly available for purchase. This story
goes on, of course, other companies have entered this market with other
devices, other services and other ‘business models’. For example, almost
all mobile phones today can do most of what an iPod can do and iTunes
today is not the only player in streaming music provision – think of Spotify
or YouTube. The lesson here is that innovation is about change and digital
infrastructures promote change and then more change.
A second interesting example is Amazon’s Kindle reader. Traditionally,
books have been sold as paper objects. Amazon bet on an alternative view:
books may be long digital messages, to be read on electronic readers.
Some advantages: lower cost, no need for space at home for storing
books, capacity for carrying a large number of books in a small device and
capacity for recovering files in case of loss. Again, we have a combination
of a new product and a new service. Not surprisingly, some publishers and
bookstores have opposed this novelty, understanding that Amazon was
taking a slice of their business. Later, Amazon has taken an even a bolder
step in starting a self-publishing service. Through Amazon, you can publish
your own books without the approval or disapproval of a publisher.
Definitively, another new business model allowed for by the pervasiveness
of digital infrastructures.
A third example is Google Search. The value of the internet would be
reduced if people did not have the means to search for and find the content
they want. Google Search offers a solution to the problem: for free, people
can use Google Search to find the content they need. Mostly impressive, the
tool works with a high level of efficiency. The massive value that billions of
users get for free cannot be calculated. Google Search multiplies the value
of the internet and the world wide web. So, what is the catch? Some say
there is no free lunch, but is this disproved in this case?
As you probably know, Google make money selling advertisements,
promising to match the advertisement as close as possible to the interest
30
of the person who is searching for something online. The business model
is quite simple: users like you and me get the benefit of free search,
but in doing our searches we give away data about our interests and
concerns. Google collects and uses this data to make money by selling
adverts targeted at the people who use their search engine. So, Google
knows that I was looking for a winter coat and I get lots of ‘coat adds’ for
a few weeks every time I go online. All three stakeholders, me, Google
and the coat advertisers win (we hope). Users like me are, however,
providing something to Google – we are ‘paying’ with the information
about ourselves. This allows Google to improve the algorithm of both their
search tool and the advertising services they sell. So far, all three parties
appear happy with this business model – we all use it!
Activity 2.6: The internet as infrastructure

In this activity, you need to research examples of products, services and businesses which
depend on using the internet as a core infrastructure. Organise three lists (products,
services, businesses) and think if these products, services and businesses would be
possible if the internet was not available as infrastructure.
Write a short report explaining how the internet has become the base for more efficient
and/or new operations. Include an explanation of how the internet emerged from an
assemblage of material artefacts, systems and organisational structures.
Activity 2.7: How have digital infrastructures enabled a business such as

TripAdvisor?
Go to the TripAdvisor main page as available in your country.
Find out and describe all the services the company provides. Investigate how the company
persuades users to contribute with content (user-generated content or UGC). What is
TripAdvisor’s business model – how do they make money from providing their service?
Who are the ‘users’ or beneficiaries of this infrastructure?
Briefly assess to what extent TripAdvisor meets the criteria to be described as an
infrastructure; check against the six characteristics in Figure 2.1). Who are the ‘users’ or
beneficiaries of this infrastructure?
2.3 Overview of chapter

A key objective of this subject guide is to prepare students to understand
how we use digital infrastructure to innovate through new digital
products, services and business models. We are going to discuss these
topics in more depth in the chapters that follow. This section presents a
summary of key ideas discussed in this chapter and how they are related
to the concept of digital infrastructures. Students can benefit from having
this understanding from the outset.
This chapter defined the concept of infrastructure. It discussed examples
of infrastructures, such as electrical power supplies or road networks
and then turned to digital infrastructures, such as the internet, the web,
cloud computing, mobile technologies and social media. The chapter gave
some examples of their impact on businesses and citizens. It emphasised
the idea that infrastructures evolve from the interplay of different layers
of technology and that previous technology influences the way new
technology evolves (path dependency). The chapter explored some
examples of new products, services and business models which have been
enabled by digital infrastructures. From these introductory concepts, the
next three chapters focus on explaining the internet.
31
2.4 Reminder of learning outcomes

Having studied this chapter and completed the Essential reading and
activities, you should now be able to:
• describe the concept of infrastructures and digital infrastructures,
providing examples
• outline the emergence of digital infrastructures and discuss their
implications for economies, businesses and society
• describe examples of products, services and business models enabled
by digital infrastructures
• explain the connections between digital infrastructure and key topics
of this course at an introductory level.
2.5 Test your knowledge and understanding

Each question below should be answered as a short essay. You should write
between 500–700 words for each answer.
1. Define infrastructures in your own words, exploring key aspects of the
concept and using examples to illustrate your points.
2. Describe in your own word the five key digital infrastructures
discussed in this chapter (the internet, the WWW, cloud computing,
mobile technologies and social media) (between 100 and 200 words
for each concept). Explain with a diagram how these are linked or
work together.
3. Give examples of how digital infrastructures enable new products,
services and business models that you use. For example, when you
listen to music, keep up with friends, pay bills or taxes, or buy clothes.
Depending on where you live of course you may do more or less of any
one thing – but for sure you use digital infrastructures for education!
32
Chapter 3: History of the internet
Chapter outline
This chapter presents an introduction to the internet, perhaps the most
important digital infrastructure in the world today. The chapter outlines
the history of the internet from its beginnings in universities and science
laboratories, its growth, first as way for scientists and academics to share
data and information and from the mid-1990s its explosive growth for
all manner of business, government and social uses. In the past 20+,
years the internet has become a universal resource at the centre of the
operations for all kinds of organisations, the lives of all kinds of people
and arguably the most effective driver of economic and social innovation
across the world.
3.1 Introduction
As you see in the paragraph above, it is easy to use the language of
superlatives when discussing the internet. It has grown in size, scope,
user base and technical performance at a fantastic rate and the speed
with which people around the world have become connected, even faster
in recent years via smartphones, is a stunning example of how rapidly
a technological innovation (a new idea and a new set of tools) can
disseminate across the world in both rich and poor countries. However,
in this chapter we will try to be a bit less excitable and set out some of
the history of the internet. In this history, we identify the key ideas that
underpin the internet as we see it today and in particular its character as
the quintessential digital infrastructure. In later chapters we discuss in
more detail some of the uses (applications) that are built on the internet –
such as the world wide web (WWW), cloud computing and social media.
In this chapter and the next we limit ourselves to the internet pure and
simple (well… not so simple).

• Define concepts of the internet.
• Give a balanced account of how the internet emerged and grew in
significance in the period from the late 1960s to the present day.
• Identify the main characteristics of the way that the internet emerged
and grew and the characteristics that have helped make it so central in
our economy and society.
In Chapter 4, other characteristics of digital infrastructures will be
presented and discussed.

• discuss the internet as an evolving concept in its historical context
• identify key steps in its evolution from a scientific endeavour to a
universal service
• describe the basic architecture of the internet, including ISPs, NAPs
and the Backbone
• introduce the domain name system.
33

[ISBN 9781292417752] Chapter 7, Sections 7.2 and 7.3. Similar content is
available in previous editions of the same book.
Ryan, J. A history of the internet and the digital future. (London: Reaktion
Books, 2013) [ISBN 9781780231129].
The book above is a very readable and chronological history of the internet
with many sources of data referenced. We particularly recommend reading
the chapters on the early internet and the world wide web.
London: Yale University Press, 2009) [ISBN 9780300151244] pp.19–35.
You can find this book online through Google Search or Google Scholar.
Cerf, V.G. (1995). Computer networking: Global infrastructure for the 21st
century.
You should also watch Vint Cerf – one of the most important engineers to
work on the internet, speak in 2009 on ‘The Evolution of the internet’ 2.38
minutes.
Watch this video on YouTube ‘History of the internet’.
Again Vint Cerf, talking this time about ICANN.

The Internet Society website has many resources to help you understand
the internet. Primary among these and at the top of their web page is their
‘Brief History of the internet’.
You can also read about the history and development of the internet on
Wikipedia. This is a link to the main entry which then links to many other
more detailed Wikipedia resources.

This chapter introduces the history of the internet including a timeline of
significant events in its early phase and introduces its basic structure or
architecture. The chapter emphasises the fundamental characteristics of
the internet as seen in various definitions.
3.2 Chapter content

The internet has a history of over 50 years. Perhaps you are not very
interested in history… but for the internet as for any infrastructure,
the way it is today is because of its history – what we have identified
in Chapter 2 as issues of the ‘installed base’ and ‘path dependency’. Put
another way, an infrastructure is its history. We will meet similar and
related ideas again in Chapter 7 when we discuss further the ideas of the
‘installed base’ and the ‘cultivation’ of an infrastructure.
The internet is not something completely new that was designed a few
years ago by some specific team of people on the basis of some specific
identified needs and implemented in a single coordinated effort. Rather,
it grew and evolved over many years and many different design activities,
starting with limited efforts to share data and computer access among
science communities in universities and laboratories in the USA, Europe
and beyond. These laboratories needed a data network to do their science,
so they tried to build one by ‘borrowing’ different existing communications
facilities and linking them together. The word ‘inter-net’ has its origins in
this earliest phase when the work was very clearly seen as being about
34
linking together existing networks that different universities and research

centres had built. These original networks were small in scale, local to
specific universities and were used mostly to move data on campus and
allow remote login to a central computer. The goal was to inter-link
these smaller networks to achieve a bigger and bigger set of connections
incorporating more and more computers at more and more sites and thus
more and more users. Today with hindsight we might call those efforts at
inter-networking ‘limited’, but at the time they were at the very cutting
edge of computer science and communications engineering.
Looking back at this history we can see that the context they faced,
needing to borrow and build on different local facilities and figuring out
how to make them work together to share data and pass messages, has
proved to be a core strength of the internet that emerged. The acceptance
of differences and the goal of doing the minimum necessary to make
diverse parts work together, explains in part why the internet has been so
successful since then at absorbing new and different technologies, new
and different users and new and different applications. The originators of
the internet were motivated by solving their own problems, rather than
pursuing business opportunities or somebody else’s problems. They were
also lucky to have resources available from national governments’ research
programmes to allow them to work freely and creatively.
The need to make something work on the basis of joining together
multiple disparate networks and the lack of commercial pressure, is at the
core of the argument that Zittrain puts forward in Chapter 2 of his book
when he discusses why, over time, competing commercial networks were
unable to survive as the internet grew. One way to state this is to say that
the internet was and still is, an infrastructure, while the competing
networks with their commercial sponsors were thought of as products or
services. The internet had a logic of ‘use me’ versus one of ‘buy me’.
3.2.1 Understanding the internet as history

As always when we discuss history, there are multiple versions of the key
events leading to today’s internet, of the key people involved and their
motives and the underlying forces that led to where we are today.
For example, some accounts of the internet place strong emphasis on the
role of the US military in developing the infrastructure, set against the
tensions of the Cold War and the possibility of a nuclear attack on any
‘centralised’ communications infrastructure. Others are more focused on
how the scientists set out to solve their own problems and along the way
created a general set of communications facilities that others could use.
Similarly, there are national biases in how the story is told – for example,
the contribution made by an early French national science network
called CYCLADES is sometimes seen as central and sometimes never even
mentioned. It depends which version you read.
The aim in this chapter is not, then, to put forward a definitive version of
how we came to have the internet of today. Rather, it is to acknowledge
that the internet emerged out of multiple efforts by multiple groups in
different counties, each with multiple motivations. If we sometimes read
somewhat contradictory accounts in different places, we do not have to
choose one over another. Rather, we can use the different emphases we
find to tell us something more general about how this infrastructure and
digital infrastructures in general, emerge and how they develop.
35
In this course and this guide, we do, however, recommend that, when
possible, you read about this history in the accounts written by the people
who were actually there at the time. In particular, Vinton Cerf, who was
a professor at Stanford University in California at the outset and a key
figure. He has written two short accounts of the origins of the internet –
one in 1995 and another (with a fellow pioneer, Robert Kahn) in 1998.
Both are referenced above in the Essential reading and you should read
both. You can also watch Vint Cerf in a video made in 2009 and cited
above.
3.2.2 Simple internet structure

At the start we need to establish some basic ideas and vocabulary to use
when discussing the internet in technical and user terms. In this guide we
will mostly speak of ‘messages’ being transmitted by the internet, meaning
in the most general terms what users want or need to be communicated.
This can mean any kind of digital data, be it numbers, text, video, voice,
be it big or small, that helps somebody perform a task. And, of course,
the internet does not care or know what the human purpose is – for the
internet a message is just a string of bits!
As we get into more detail, we will talk about these ‘strings of bits’ in
terms of other units of data:
A bit – a binary digit (a 0 or a 1), enough to encode a yes or a no, a true
or a false.
A byte – 8 binary digits, this is enough to encode one text character in the
Western alphabet. One byte can hold 28 different patterns of bits (codes) –
e.g. in decimal from 0 to 255.
A word – 4 bytes or 32 binary digits. This allows 232 codes. Most internet
addresses (IP addresses) used for a sender and a receiver of a message are
one-word (32 bits) long. It means that we can have over four thousand
million different addresses for computers (hosts) on the internet (but it
turns out that this is not enough separate addresses – see Chapter 5 and
the discussion of IPV6).
A packet – a bundle of user data plus the address of the sender and of the
receiver and other control information. We will talk about packets when
we discuss packet switching and the IP that the internet uses to route a
message from sender to receiver.
Consider Figure 3.1 below – a very simplified model of the internet. This
shows a number of computers – called Hosts and shown as hexagons
– that are each connected to some national or regional networks (four
in the diagram – remember it is the ’inter-net’). Hosts can be big – like a
university computing centre or a big company, medium-sized like a single
supermarket store, or small like my smartphone. At the top right of the
diagram is a cluster of small hexagon-Hosts to represent smartphones.
Each host is connected to at least one network, some (the bigger ones)
are connected to more than one for redundancy and resilience. For the
network to work (e.g. reliably deliver packets of data) each Host must
have a unique and unambiguous address.
36
Figure 3.1: A simple schematic diagram of the internet

Each network is shown as a set of Nodes (circles with an N). The nodes
are linked to each other by some communications medium – probably
cables (copper or fibre optics), or perhaps radio or satellite links. Whatever
the physical medium of the links, they all allow data (messages) to be
passed. Nodes are the switching points in the network and they are
also called Routers. As messages travel across this network in packets,
each individual node where they arrive will decide which link to use to
send the message on its next leg (or next hop), based on the destination
address. Conceptually this is exactly what a person employed as a post
office sorter does (though today it is probably a machine). Person or
machine, what they do is to look at the address of a packet or letter and
put it in the appropriate bag for its onward journey. A postal letter would
expect to be sorted in this way a number of times before it is delivered.
In the centre of the figure are nodes marked with a B. These represent the
backbone of the network – the place where high volumes of messages
will flow between different networks and which need higher capacity in
their links and switches. The backbone links are shown with a thicker line
to indicate their higher speed and capacity.
Consider a message from Host A (top right) to Host Z (Bottom left). What
route will it take and how many hops will this include (a hop is from one
node to another)? Well, we cannot really give a definitive answer since
there are many possible routes. It will depend on how each node chooses to
route the message and the best way may depend on how much other traffic
is using the network. Just like driving in a city – at different time of day you
may well take different routes. We call this approach dynamic routing.
Activity 3.1: Developing definitions

From this very simple diagram and the accompanying explanation, we have introduced
some key ideas and new vocabulary specific to the internet e.g. the words in Bold. Using
your textbooks and online resources, confirm your understanding of each of these terms
and write a brief definition for each in your own words. Do not just copy and paste from
the internet – engage your brain in the activity.
37
3.2.3 Defining the internet

Figure 3.1 above gives a simple schematic model of the internet in a
diagrammatic picture and names for the various parts. Now we can be a
bit more specific and consider how we might describe the internet itself in
a simple sentence or two. A brief description in our own words could be as
follows:
The internet is a shared public facility for sending digital data
from one place in the world to another in a fairly reliable way.
This facility to transmit data is open to use by all and can be
used to serve many purposes. All users receive equal treatment
for their data as it is transmitted.
In Laudon and Laudon, Glossary at the end of the book, they define the
internet in even simpler terms as:
‘Global network of networks using universal standards to
connect millions of different networks.’
The key elements of these two candidate definitions can be teased out a
bit and include:
1. transmits digital data – it does just the one task of sending digital data
(e.g. anything coded in 0s and 1s)
2. from one place to another – data goes from point to point so we need
consistent ‘addresses’ for these places so the data can be transmitted
3. network of networks – it is made up of multiple different
communications facilities all linked together and adhering to some
minimum common standards
4. global reach – it is big and operates (almost) everywhere, e.g. via
smartphones, satellites etc.
5. shared and public – it is available to all (billions of users) who share
the network resources
6. fairly reliable – the policy of the internet is described as ‘best effort’
and we cannot rely 100% on it so we need some backup procedures,
security measures etc.
7. equal treatment of all data – once data is in transit your data gets
equal treatment to my data (but we discuss this issue more under the
theme of net neutrality in Chapter 16).
Activity 3.2: Data transmission across the internet

Thinking about these seven key themes and reflecting that the internet is a
communications facility, make specific comparisons of each theme with the way that the
global postal service works. Of course, the postal service is another, older, communications
infrastructure and is not in all respects quite like the internet.
When doing this, consider that the postal service too is available to all, takes letters
(messages) from one place to another around the world using a system of addresses and
is made up of various national postal services that collaborate to send a message. It is
fairly reliable but not 100% and it gives equal treatment to most letters.
A more elaborate and technical definition of the internet from a US

government organisation written 20 years ago is available via the Federal
Networking Council Wikipedia entry (the definition is also cited in Cerf,
1995):
38
October 24, 1995, Resolution of the U.S. Federal Networking Council

RESOLUTION:
The Federal Networking Council (FNC) agrees that the following language reflects our
definition of the term “internet”.
“internet” refers to the global information system that:
(i) is logically linked together by a globally unique address space based on the internet
Protocol (IP) or its subsequent extensions/follow-ons;
(ii) is able to support communications using the Transmission Control Protocol/internet
Protocol (TCP/IP) suite or its subsequent extensions/follow-ons and/or other IP-
compatible protocols; and
(iii) provides, uses or makes accessible, either publicly or privately, high level services
layered on the communications and related infrastructure described herein.”
Available at https://www.nitrd.gov/fnc/internet_res.aspx.
This is a more formal definition – but essentially says the same things
e.g. the internet is a global system for communications and can serve all
manner of uses ‘layered on it’ as they say and reflecting a layered model
(see Chapter 6 for more discussion of layered models). This definition also
stresses the central importance of the common rules of operation that the
internet uses – known as protocols. Interestingly, it also shows us that
the people framing this definition 20+ years ago were fully aware that the
internet would continue to develop over time and so they use the phrase
‘…or its subsequent extensions/follow-ons…’ in their definition when
discussing the core protocols (TCP and IP).
The protocols they identified in 1995 as central are still central today.
These are Transmission Control Protocol (always called TCP) and the
internet Protocol (always called IP). These protocols are considered in
greater depth in Chapter 5, but for the moment we should understand
that:
• IP is concerned with the rules for addresses and for using them to get a
message ‘across’ the internet from its origin to its destination on a ‘best
effort’ basis.
• TCP is concerned with how messages are reliably delivered from a
source to a destination and what to do if a part of a message is lost
along the way (it does happen).
Activity 3.3: Conceptualising the internet

Find two or three more definitions or models of the internet, in books or online – for
example in dictionaries, Wikipedia or other textbooks. Consider these definitions and
describe the extra aspects or characteristics of the internet that they offer or emphasise.
3.2.4 Internet evolution: from a scientific endeavour to a universal

service
In this section we provide a brief overview of the evolution of the internet
in its early years. It is presented as a timeline. The events shown are based
mostly on Zittrain’s book (Chapter 2, p.27 onward), Cerf’s brief paper on
the origins of the internet (1995) and the timeline on the Internet Society
website. You should read Zittrain’s chapter 2 and Cerf’s (1995) paper, as
well as this section, before attempting the exercises.
39
Brief timeline of the internet in its early years: 1969–1998

1969 US-funded ARPANET project launched: first message sent between
computers at UCLA and Stanford universities in California. By end of the
year 4 universities (hosts) are linked, adding University of California Santa
Barbara and University of Utah
1973 First international host connection is made to UCL in London
1974 First versions of the protocol TCP/IP is presented
1978 IP separated from TCP so they can develop independently
1983 ARPANET moves to using TCP/IP as its core protocols
1984 The domain name system (DNS) introduced
1988 First chat programs (IRC) started to be used
1989 Number of hosts >100,000
1990 First commercial dial-up link to the internet in the USA (commercial
consumer ISP)
1990 First world wide web site at CERN
1992 Number of hosts >1M
1995 Main backbone passed from US government to commercial suppliers;
Hosts > 5M
1998 IP version 6 (IPV6) launched – allowing for many more host addresses
Table 3.1: internet timeline
Ever since computers were first used in the late 1940s, people wanted to
network them, that is both to send data between computers and for people
to access computers remotely (and remote could mean down the corridor
or around the world). Up until the late 1960s this was either a very local
affair (terminals connected to one computer in a single building), or vastly
complex and based on proprietary technologies. There were no general
networks available for average users to send and receive data, other than
at very slow speeds using the telephone network and a device called a
‘modem’. Still, scientists and researchers (and their funding bodies) were
becoming big computer users and were very interested in sharing data
and providing a communications infrastructure in support of science and
technology development programs. Governments in the USA and Europe
sponsored national initiatives to build such networks, as did the US military.
So, many research projects and small-scale experiments took place.
One in particular is usually identified as the internet’s principal origin
– this is ARPANET, a project of the US Defence Department Advanced
Projects Agency (DARPA) – initiated in 1967 and starting to operate in
1968. The ideas and technology embodied in ARPANET demonstrated
themselves to be superior in general to other initiatives, as a way to
connect heterogeneous computers and local and regional networks
together (heterogeneous = all many kinds, a mix). Zittrain (2008, p.28)
describes it admiringly as a way to ‘cobble together existing research and
government networks and wring as much use as possible from them’.
Over the next 15 years or so most other networking projects were eclipsed
or became integrated in to the ‘internet’. Across the world scientific
researchers and other technology users and network managers started
to use network services connected to the internet and using the internet
protocols (TCP/IP) and other internet services. These included early
versions of shared files, email, chat programs and, later on, the world wide
web (WWW) with its web pages, browsers and search engines.
40
Zittrain also notes that the motives or philosophy of the ARPANET/

INTERNET developers was distinct and in particular very different from
the competing commercial network companies which existed at the time
and who sought to exercise control over how their networks were used and
make revenues from their users. He quotes the internet pioneers as saying,
‘We reject kings, presidents and voting. We believe in: rough consensus and
running code’ (Zittrain, 2008, p.28). This is a philosophy that still underpins
the internet: the internet still has no single, central, formal management
structure (see also Laudon and Laudon, Chapter 7, Section 7.3). It is also
related to the philosophy that is seen in open-source software communities
and can be described as an agile approach.
By the early 1990s, the internet as we see it today was well established
in academic and research contexts. Universities for example were among
the first large organisations that gave all their staff and students email
addresses. The internet was also attracting more and more interest from
commercial and business users. In theory, commercial (for profit) use was
banned, but still it started to occur. To regularise the situation, the internet
was separated from the US government and the US government research
funding that had kept the internet going was withdrawn. To achieve this the
central management of the network traffic (the backbone) was redesigned
and then contracted to a number of competing commercial companies (see
Cerf, 1995).
From 1991 the whole world was welcome on the internet and the millions
of people with a home computer could connect via an internet Service
Provider company (ISP) and benefit from web searches, email and – slowly
at first – e-commerce. At ha time there was a rather limited service for
individuals, still using dial up modems and telephone lines, but by the late
1990s in the developed world almost any home computer could have full
internet connectivity using a much faster broadband connection, although
usually still via the landline telephone cables (a system known as DSL
Digital Subscriber Line), but without interfering with the normal telephone
usage. For many people this remains the technology in use today. For others,
a broadband connection is made via cable television networks and in both
cases these links are moving from copper cables to fibre optic cables for
greater speeds. Other people in rural areas may today be using a satellite
connection or link via mobile phone network. This diversity of access once
again emphasises the internets ability to evolve and draw on multiple kinds
of network technology.
And of course, very locally in your house, school, factory or hospital,
connections to smartphones and mobile devices are today made via WiFi
linked to a local wireless router – a kind of mini Node. To offer even more
choice of how to connect (and make it even more complex) a smartphone
can also, when away from WiFi, make its link to the internet via the mobile
phone company’s data services. Some people rely wholly on their 4G/5G
mobile phone data connection for their link to the internet.
It is not our purpose in this course to go any deeper into all these local
connection technologies – what is often called ‘the last mile’ – in technical
terms. But it is important to recognise that the multiple ways that we
connect to the internet as individuals and how businesses and organisations
connect too, are all based on a set of standards and protocols that ensure
that independently of the way we connect, we can all use the same digital
infrastructure whatever our device or machine. From the largest corporation
down to a simple tablet or a network-connected security video camera, we
can ‘speak the same language’, or perhaps a better metaphor, we can all ride
on the same roads and obey the same traffic rules.
41
Activity 3.4: Communication with the internet

In a normal week, how many different ways do you use to connect to the internet? How
many devices do you use to connect to the internet? Include what you do at home, with
friends and at work or college. You may need to think about your television, phone and
computers but also other devices such as your car (satnav?), washing machine (service
record?), cameras, credit cards(?) travel tickets(?) and so on. Try to match each device or
kind of use to some ‘last mile’ technologies such as 3G/4G, Bluetooth (last metre), WiFi,
NFC, RFID, DSL and so on. Try to make your list as long as you can even if you are unsure
of the actual technical details.
3.2.5 Architecture of the internet

While digital infrastructures like the internet are not usually the result
of a single project or a single design, but rather evolve over time, they
usually have a coherent structure that they retain as they evolve. This is
what we refer to as their ‘architecture’. So, from the top-down perspective
and sustaining it since its origins, the internet’s architecture is one of a
network of networks (see also Figure 3.1). But in order for multiple
networks (measured in 10s of millions today!) to operate together and pass
messages between them they need some regular structure and ground rules
for their interconnection.
At the highest, level is the Backbone, which is where messages are
exchanged between major sections of the internet. The backbone has
to have a lot of capacity to exchange data, working a bit like a six-lane
highway with a very high-speed limit. The backbone, as a part of the
internet, is itself a ‘network of networks’. When it was passed into the
hands of private companies in the 1990s it was deliberately done in a way
such that no one company could control it. Today’s internet backbone
connections have a speed (basic capacity to shift bits) of 2,488 Mbit/s or
faster (Mbits/s = 1000 bits per second) and they use very fast switches
(routers) and high-capacity fibre optic cables. The core backbone is still in
the United States and other regions and countries have their own backbone
networks e.g. in Europe, Asia etc. (see the Wikipedia entry).
At the lowest level, are the individuals like us, with our home computer
or smartphones connected to an ISP or a college campus network (e.g. by
WiFi as described above).
In between the backbone and an ISP are a rich tapestry of local and
regional networks described as MAE (Metropolitan Area Exchanges) and
Regional Hosts in Laudon and Laudon (2021). Of course, not all internet
traffic will need to through the backbone. If I send an email to another
UK person using a UK based email service, then it can be directed straight
there within the UK at some lesser internet exchange point or MAE, or it
may transfer at the level of the European Backbone. See the Wikipedia
entry on internet exchange point.
So, the route for an email sent by me (in London) to you (let’s say in
Japan, Mozambique or Australia), will be made from my phone to my ISP
(perhaps my UK mobile company, perhaps the university WiFi and MAE
connection), from there to a regional network in the UK, then up to the
backbone. Then the message will traverse a similar set of layers from the
backbone to, for example, a Japanese national network and then on to
an ISP to deliver the email to your inbox wherever you email is hosted
(perhaps on Gmail in California even though you live in Australia). (The
digital infrastructures do crazy things with geography – you can live in
one continent, keep your email in another, buy clothes from a third and
42
keep your bank account in a fourth. Sometimes people talk of digital

infrastructures as leading to the ‘death of distance’).
The obvious next question to ask and which we address in the next two
chapters, is how does my message actually make its way through such a
complex set of interconnections (and do so in seconds)? The answer is
founded in the inspired choice made over 30 years ago by a small group of
engineers to build a digital inter-network (we would better say a digital
infrastructure) using packet switching and designing for this the TCP
and IP. This is what we address in the next chapter.
Activity 3.5: Mapping the internet

Using the internet as a source of information, prepares a map of how your country links
to the internet and show the organisations that own or control the relevant networks
(e.g. phone companies, ISPs, Exchanges and Backbone connections).
To start, try a search for e.g. ‘Malaysia internet backbone’ or ‘Hong Kong internet
backbone’ or perhaps ‘Australia internet exchange’.
Do both a general search and at an ‘images’ search. The images search should reveal
maps and diagrams. Try not to be put off by all the technical jargon and abbreviations you
find (we know that you are probably not an electronics or communications engineer and
do not want to become one). But use this information and maps to get the bigger picture.

This chapter introduced the internet as the prime example of a digital
infrastructure in our era. It explains some of the history of the internet
and some of the core aspects of how it is shaped and organised (its
architecture). The chapter also introduces some definitions of the internet
and thereby shows some of its core characteristics.

• discuss the internet as an evolving concept in its historical context
• identify key steps in its evolution from a scientific endeavour to a
universal service
• describe the basic architecture of the internet, including ISPs, NAPs
and the Backbone
• introduce the domain name system.

1. Explain the origins of the internet and the goals of the pioneers who
worked on it. Show how their work and the decisions they made have
influenced the evolution of the internet we use today.
2. Provide your own definition of the internet in your own words and
appropriate to today. In your definition you should include reference to
the internet’s scale, the variety of uses it enables and the consequences
of its use.
3. Write a short essay to introduce the concept of an infrastructure having
an ‘architecture’. What do we mean or imply by using this word?
43
Notes
44
Chapter 4: Emergent design of the internet
Chapter 4: Emergent design of the

internet
Chapter outline
This chapter explains the most important features of the design of the
internet, in a basic level. The idea is to present how the internet transmits
data across Hosts. It explains in more detail the use of packet switching
for transmitting data, which gives the internet both the capacity of having
scalability (the volume of data transmission can vary substantially) and
resilience. The chapter also introduces the domain name system (DNS)
and the URL-style addresses (Universal Resource Locator), explaining how
content is found on the internet.
4.1 Introduction
The internet has been a phenomenal success as a digital infrastructure –
an infrastructure for everybody and for the whole world. So, the obvious
question to ask is ‘Why?’.
After all, as Zittrain (2009) explains in Chapter 2 of his book, there
were plenty of competitor projects and local or national networks under
development and many of these had more substantial financial and
commercial backing than the pioneers working on the internet. In this
chapter, we look at some of the key decisions that this team made about
how to structure and operate the internet and which we can argue led
to its success. We will argue that it was in large part these decisions that
laid the foundations for the internet’s ability to scale up to be a global
infrastructure serving the needs of billions of users and transmitting
unbelievable qualities of data.
In this chapter we focus on three specific technical decisions that emerged
in the early days and that have been proven over the years. In brief these
are to:
1. transmit data: do just this one thing and let all other things, any
intelligence or value adding activity, occur at the network fringes, e.g.
in the Hosts
2. use packet switching to optimise capacity and help build in scalability
and levels of resilience
3. design a flexible addressing system able to scale up from less than 10
hosts to millions (even billions) of Hosts.

The aim of this chapter is to identify and justify some key characteristics of
the internet that have helped to make it so successful.

• describe the benefits of the chosen packet switching approach
• explain how the internet addressing system works
• explain how these specific issues relate to digital infrastructures in
general.
45

[ISBN 9781292417752] Chapter 7, Sections 7.2 and 7.3. Similar content is
available in previous editions of the same book. Read the chapter in other
editions that discusses infrastructure.
You may find a copy of this book online through Google Search or Google
Scholar.
Cerf, V.G. ‘Computer networking: global infrastructure for the 21st century’
(1995).
Kahn, R.E. and V.G. Cert ‘What is the internet (and what makes it work)’
(1999).
You should also watch Vinton Cerf, one of the most important engineers to
work on the internet, speak on ‘The evolution of the internet’ (2009).
Video: DNS as fast as possible (2016).

In this chapter we look at three key decisions made as the internet was
being designed and established and in its early history. Looking back
with hindsight these can be seen as key to the internet’s success. It is also
argued here that they can also be used as exemplars of more general
characteristics of all other digital infrastructures.
4.2 Chapter content

The three decisions explored in this chapter are: first to be a ‘dumb’
network that does one thing well, second to use packet switching and third
to develop a flexible and dynamic addressing system. We devote a section
to each one.
4.2.1 Do one thing well – leave other things for other people
At the outset in the 1980s computer scientists such as Vinton Cerf were
working hard to link together (inter-connect) big mainframe computers
located in universities and research centres. These machines were built
by different manufacturers, did diverse kinds of work and to link them
together would need some kind of general purpose and adaptable solution
that could incorporate more computers and more variety in their design
and what they were used for – not one optimised to the specific needs of
the moment. It was the case then (even more than today perhaps?), that
computers were developing very rapidly, so in five years’ time they knew
that there would be different machines doing different jobs, but they
should still be able to use the intercommunication system.
The approach taken, given this context, was to develop a general-purpose
data communications capability, but to isolate it from any specific machine
by using a network access device called an IMP (Interface Message
Processor; read Cerf, 1995). The IMP developed at the time was a small
computer in its own right that could ‘hide’ any computer and its specific
details from the network and the network from the computer. To add a
different or new computer to the network the only work needed would be
to reprogram the IMP.
In terms of Figure 3.1 the IMP was placed between a host and the link to a
node of the network. It isolated the one from the other and allowed both
‘sides’ to develop and change without affecting each other. Indeed, the IMP
from the 1970s has evolved to be the IP.
46
One way to express this is to say that the internet was (and still is) a
deliberately ‘dumb’ network. It needs to know almost nothing about who
sends or receives data – only the bit string that is the message and the
addresses of sender and receiver. The internet does not care what the data
represents – it was and still is just a string of bits – and it undertakes to
do nothing clever with the data while it is in transit. Also, it did not (and
still does not) keep any kind of detailed billing data to count bits and send
out bills (unlike your mobile phone network!). Today the service providers
manage the access to the internet, billing final users (individuals,
companies, governments, etc.). Remember though that at the outset in the
1970s and 80s the internet was subsidised by the US government and so
did not need to account for specific usage in financial terms.
The genius of the pioneers was to recognise that building a dumb network
was easier than a smart or busy-body network and (more importantly)
helped to make a network architecture that could scale up (e.g. from four
Hosts to millions or billions). This has also meant that any intelligence or
extra features has always had to be added on the outside – that is in the
Hosts not in the network. This decision has proved very powerful as a way
to drive innovation based on using the internet. A person with a good
idea – let’s say a streaming service for films with real time AI-based audio
subtitling in any one of 300 languages (note this is a made-up example) –
can program their own computers (Hosts) to do the clever stuff and let the
‘dumb’ internet carry the service to their customers across the world. The
network can accept this innovative service just as it can accept thousands
of others.
4.2.2 Use packet switching to share bandwidth and build in

resilience
The internet is what is called a packet-switched network. That is,
messages (data to be transmitted) are broken up into separate chunks –
we call them packets – and each packet is sent independently over the
network. The file for this guide in Word format on my computer is about
1,100 KB of data, if I sent it to my fellow author using the internet it will
not be sent as one big chunk of data, but be split up into about 1,100
separate packets, given that an internet packet for the IP is on average
around 1,000 bytes (e.g. 1,000 characters). (The actual maximum of data
in a packet is 64,000 bytes, but for good reasons such large packets are
seldom used – see the next chapter).
The IP as the successor to the original IMP will chop up the file and
generate these packets. Each packet will, in addition to a chunk of the file
(a string of bits often called the payload of the packet), contain about
24 bytes of data for the sender’s address and the destination address and
other control data including a sequence number to help reassemble the file
at the other end.
When the file is sent (e.g. from Tony to Magda) packets will travel in
both directions, bigger packets carrying data to the destination (Magda’s
computer) and smaller packets as acknowledgements sent back to
the sender for each arrived packet (Tony’s computer). So, in practice,
many packets traveling on the internet will be very short – just an
acknowledgement of delivery perhaps. Even when sending data, it is not
usually a good idea to let packets be too big. (Note all the numbers used
here are rough estimates in decimal numbers – for finer detail you would
need to look up packet size distribution in a technical reference. However,
for this course it is the principles that we need to learn – not all the detail.)
47
When Tony sends this file, he should not expect that these 1,000 packets
to all travel along the same route, arrive in the order sent, or even arrive
at all. But do not worry, between them TCP and IP – the core internet
protocols – will sort this out: what we call the ‘reassembly’ of the original
stream of bits from the many packets arriving at the destination.
You might very well ask, why go to the bother of slicing up a message into
packets and sending them independently through the network? It perhaps
does not sound very efficient or reliable. In the early days the developers
of the internet were in a minority in choosing this approach and many
thought it inappropriate. The alternative was a circuit switched
network, as used in older data networks and in the original telephone
systems. In a circuit switched network the first task is to establish a circuit
from one end to the other (e.g. dial a number and wait for the telephone
exchange to go click-click-click as it sets up a solid copper wire connection
from caller to receiver). Remember, up to the 1950s telephone exchanges
were huge electro-mechanical devices that used physical switches to set up
a line and make a call possible. Once that connection is made, the data (a
conversation) can flow, but not before.
However, using packet switching for digital networks is now almost
universally adopted. Put simply, if a network is used efficiently as a shared
resource (a core characteristic of an infrastructure), the primary concern
is with the best use of the overall system’s capacity to move data. By
breaking messages down into packets each node in the network, and each
link between nodes, can be kept busy handling packets without needing
to bother that these packets are part of the same message – they are just
packets of data to be sent on their way. Keeping each link busy (e.g. high
utilisation) equates to efficient use of the network. This is sometimes
explained in terms of these networks being designed for ‘bursty’ traffic
– that is streams of data that include bursts of intense activity and then
periods of little or no data to transmit (silence). In circuit-switched
networks, resources (e.g. links) are held for just one message to use. But
with packet-switching resources (e.g. links) do not need to be held for
periods of silence.
The technical term for this sharing of a link by different and distinct
streams of data is ‘multiplexing’, sometimes called ‘muxing’. It can be
defined as sending multiple streams of data over a communications link at
the same time in the form of a single, complex signal. At the receiving end
the separate streams of data are recovered (separated out). This process
is called demultiplexing or demuxing. As a simple analogy, a post office
truck full of letters and parcels driving up the highway from London to
Edinburgh is a multiplexed link and the sorting office in Edinburgh will do
the demuxing.
4.2.3 A dynamic domain name system (DNS)

The third core decision in the early days of the internet was to design a
very flexible way of giving Hosts unique addresses. First of all, the early
decision to use 32-bit words for IP addresses foresaw the potential for
many Hosts in a network. This type of address (IP address) is still used
today in IP version 4 (IPv4) – the most commonly used version. However,
in IP version 6 (IPv6) the size of addresses is increased from 32 bits to 128
bits giving a staggering number of possible addresses – IPv6 and details of
addresses is discussed more in Chapter 5.
You and I usually type an internet address in a textual form, for example
we type www.london.ac.uk when we use our browser. Similarly, we send
an email to [email protected]. We call these URL (Universal Resource
48
Locator) style addresses. This way of writing the address suits us as

humans, but our computers have to find the actual IP address (the 32-bit
number) that these refer to. To do this, the internet needs some kind of
‘telephone directory’ to look up a name ‘www.lse.ac.uk’ and get back the IP
address as a number – usually written in the form 212.113.11.22 which is
called dotted decimal notation. Remember an IP address is 32 bits =
4 bytes, each of which can hold a number from 0 to 255. So, this notation
uses 4 decimal numbers to stand for the 32-bit binary number. So, we
could convert 212.113.11.22 into binary as 11010100.01110001.00001
011.00010110. (This conversion was done using a decimal to binary tool
found using Google.)
To better understand IP addresses and URLs we also need to understand
IP address domains. These are familiar to us as the last part of a URL:
.com .info .net .uk .nz .edu .cn .sg .jp are all top-level domains.
For the purposes of addresses each domain relates to different parts of
the internet – and they may be geographic (.uk .fr .cn) or they may be
thematic (.com .info .gov .coop). Over time the internet has set up and
authorised more and more such domains - the list is now quite long (see
List of Internet top-level domains, Wikipedia).
So, if we look at www.london.ac.uk, we can see it is in the .uk domain.
The middle of the URL is then a sub-domain of .uk ‘london.ac’. The start
of the URL is an optional subdomain prefix – in this case www meaning
world wide web. In practice you do not usually have to use the www bit –
it is the default destination for any request initiated from a browser.
An IP address is needed to send any message via the internet – and is 32
bits (IPv4). However, just as a URL – some of those bits are used to specify
the domain and some to specify the Host address within the domain.
We call these two parts the network ID and the Host ID. The network
address comes first (even though the domain comes last in the URL) and is
the most important part for routing the message.
The actual scheme used is able to handle a few domains with millions of
Hosts (.com or .uk or .net), as well as other domains with far fewer Hosts
(.mq Martinique or .gw Guinea-Bissau). For the first type just one byte or
less is used for the domain’s network ID, while for the latter it may be 3
or more bytes. The actual working of this scheme is quite complex and
beyond this syllabus. But if you want to know more, see Classless Inter-
Domain Routing, Wikipedia.
4.2.4 The DNS in use

However, an IP address is structured, it will usually need to be found (e.g.
looked up) before any message can be sent. Doing this is the job of the
Domain Name System (DNS) which provides IP addresses by their
billions every day. It is core to the running of the internet and itself it runs
on the internet. When it fails, we certainly notice (see below). Thus, your
computer or smartphone uses DNS to ‘look up’ the URL-style address every
time to find the equivalent IP address.
One big advantage of having the DNS available online and looking
addresses up every time (almost), is that it can be kept up to date. So, if you
change your company’s name and web URL from www.xxx.com to www.
yyy_and_zzz.com, or if you change your IP address because you change the
ISP where your website is located, you just need to update the directory
(DNS) which everybody uses and they will still be able to find you.
49
local
Mail client cache
local
cache
DNS Resolver
recursive
Web Browser DNS Resolver DNS
cache timeout: mini search
1.30 min cache Operating System Your ISP
Client Programs
Your Computer
Figure 4.1: Using the DNS.

Source: Picture copied from https://commons.wikimedia.org/wiki/File:DNS_in_
the_real_world.svg
As Figure 4.1 shows, each Host on the internet can keep a small list of
IP addresses in a local ‘cache’ (note that a cache is the name computer
scientists give to a temporary set of data held locally to speed up a
process). If the cache cannot help translate the URL to an IP address, then
the request is passed to your ‘local’ DNS. This is probably set by your ISP
or the similar organisation which provides your internet services. But
you can choose to use an alternative DNS server such as that offered by
Google. If your local DNS server does not know the IP address that goes
with an address like xyzabc124.com.zn (which is not surprising because I
just made it up), then it in turn can ask a bigger global DNS which should
know, or know who does know (assuming it is a valid internet address).
In this case and since the domain .zn does not exist as an internet domain,
we will never get a valid IP address back from the DNS however many
servers it consults. But if that were .nz and not .zn, then your local DNS
will try to answer the question by passing it to a global DNS which will
pass the question to the .nz DNS in New Zealand – what is known as the
authoritative name server for that domain. This passing of requests
across the internet to find a DNS that ‘knows’ is called recursive DNS
search.
Activity 4.1: Understanding internet addresses

Using you own browser type in these two addresses:
xyzabc124.com.zn (an impossible internet address – there is no domain.zn)
and
xyzabc124.com.nz (a possible internet address, but only the .nz domain (New Zealand)
would know if it is valid or not)
In each case note the message that comes back. Then explain why these messages are
different and what this implies as to how they are handled. Why are they different?
Now watch this video ‘DNS as fast as possible’ to deepen your

understanding of DNS.
4.2.5 What if DNS fails?

At about the same time that this section of this subject guide was first
being written, the world saw the result of a failure of the DNS. On 21
October 2016 DYN (dyn.com), a company that offers DNS services to
many US-based web users on the east coast, experienced a failure of their
system. The cause was a DDoS (Distributed Denials of Service) attack on
the company by persons unknown. In a DDoS a coordinated attack is made
50
by sending a huge number of requests, updates or amendments to a Host

such that computers are overwhelmed. In this case the messages came
from all manner of internet devices such as cameras and video recorders
that had been ‘hacked’ and were running malware under the control of
the attack’s initiators (botnets). The result of no DNS was that many big-
name web businesses were out of action, including Amazon.com. Read
more about this attack on the BBC website: ‘Smart’ home devices used as
weapons in website attack.
Dyn is a specialist company that delivers web services, and they were
able to act to blunt the attack and return its service to normal. Executive
Vice President, Scott Hilton said: ‘This attack has opened up an important
conversation about internet security and volatility. Not only has it
highlighted vulnerabilities in the security of “Internet of Things” devices
that need to be addressed, but it has also sparked further dialogue in
the internet infrastructure community about the future of the internet.’
(Oracle.com).

The chapter analysed the three key decisions made as the internet was
being designed, which can explain the internet’s success. The chapter
explained the most important features of the design of the internet,
in a basic level, presenting how the internet transmits data across
Hosts. Particular attention was given to the use of packet switching for
transmitting data, which gives the internet both the capacity of having
scalability and resilience. The chapter introduced the DNS and the URL-
style addresses.

• describe the benefits of the chosen packet switching approach.
• explain how the internet addressing system works.
• explain how these specific issues relate to digital infrastructures in
general.

1. Why has the internet been so successful over the years? In your answer
focus on two specific features that you think lie at the centre of its
success.
2. Explain the technology of packet switching and its advantages for
digital infrastructures. Why is packet switching important in explaining
the internet’s ability to scale up from its modest beginnings to a point
where it serves huge numbers of users and huge numbers of Hosts
across a large number of networks?
3. Draw a poster, diagram or flow chart to explain the way that people
write internet addresses and then how they are translated by the DNS
system into IP addresses that can be used in data packets.
51
Notes
52
Chapter 5: The internet today and the TCP/IP
Chapter 5: The internet today and the

TCP/IP
Chapter outline
This chapter is concerned with the rules that the internet uses to ensure that
messages are transmitted from sender to receiver. These rules, known as
protocols, are programmed into computers of various kinds (Hosts, Nodes,
etc.). The protocols we discuss here are at the core of the internet. They
are the internet protocol (IP), the Transmission Control Protocol (TCP) and
User Datagram Protocol (UD). Most of the chapter relates to the first two.
UDP is a simpler protocol and we have less to say about it. Nonetheless it is
an important part of the internet’s protocols.
5.1 Introduction
This chapter is concerned with how the internet main protocols work and
the data structures they use (e.g. packets in IP and data segments in TCP).
You can think of each protocol as a set of rules of behaviours for devices
connected to and working as part of the internet. As in earlier chapters, the
aim here is for you to understand the jobs that these protocols do and the
principles that lie behind their design. We are not too concerned with the
finer technical details. Then, of course, computers (Hosts, Nodes Routers
etc.) need to be programmed to obey these rules and that can raise many
technical concerns – but here we are trying to understand the principles and
understand how messages pass across the internet.

The aim of this chapter is to introduce the concept of a network protocol
and to develop a logical understanding of the internet’s three primary
protocols: IP, TCP and UDP. To do so the chapter introduces some specialist
vocabulary associated with digital networking.

• explain the character and purpose of a network protocol in general
• give a logical description of the specific tasks that IP, TCP and UDP
support as part of the internet suite of protocols
• understand the main developments that arise as the internet moves
from IP version 4 (IPv4) to IP version 6 (IPv6).

[ISBN 9781292417752]. Chapter 7, Sections 7.2 and 7.3. Similar content is
available in previous editions of the same book. Read the chapter in other
editions that discusses infrastructure.

See also Linus Sebastian’s video on ‘What is TCP/IP?’. Probably best
viewed after reading the chapter since this video introduces the topics in
the opposite order to this chapter. It also has adverts at the end which you
should skip.
53

This chapter introduces the concept of a network protocol. It considers in
detail three key protocols used by the internet – IP, TCP and UDP – which
have been introduced briefly in previous chapters.
5.2 Chapter content

So far, we have discussed packets and addresses, but not the way that
the internet gets a message reliably from Host A to Host B. We know that
packets containing data are generated by the sender and then passed to
their local network node (router or switch). And we know that it is then
the responsibility of that node to decide where to send the packet next –
e.g. to route it, hence we call these switches routers.
In this chapter we will look at the details of how this happens and, in
particular, the role of the internet protocols IP, TCP and UDP. In Chapter 3
we introduced TCP and IP as fundamental to the origins of the internet and
a key part of its success. Then, in Chapter 4, we related the IP in particular
to the need for an addressing system, the idea of using packet switching and
the DNS. These three protocols are representative of the central layer of a
digital infrastructure – what we called the protocol layer. See the simple
three-layered model (Figure 5.1). Now we can look a bit deeper into the
actual rules that these protocols propose for managing message flows and
the data structures that they use (e.g. IP packets and TCP Segments).
Host A running Host B running

an application an application
Protocol TCP is
host to host or
Application Application end-to-end
TCP IP works between
Protocol IP IP Protocol hosts and nodes and
Network
Network between networks
Physical Physical
Network
Internet shown here
Network
as network of networks
Figure 5.1: The internet as a simple set of layers.
5.2.1 What is a protocol?

First, however, we want to explore the idea of a network protocol a bit
more.
In the most general sense, the word ‘protocol’ means a guiding set of rules.
The Oxford English Dictionary says, among many other definitions, ‘the
accepted or established code of behaviour in any group, organisation, or
situation; an instance of this’. In other words, an agreed set of rules for
doing something.
In the same spirit they also give a definition specific to computing and
telecommunication and rather more relevant to our concerns: ‘A (usually
standardised) set of rules governing the exchange of data between given
devices, or the transmission of data via a given communications channel’.
So, a protocol is a set of rules, standards and behaviours and can be
applied specifically to exchange of data – which is the business of the
internet or more generally of digital infrastructures.
The protocols we consider here, IP, TCP and UDP are certainly
standardised across the world. These standards specify rules that are
54
programmed into computers (Hosts, Nodes, Routers, mobile phones, TVs

and dishwashers – indeed any device with an internet connection etc.) and
the devices will obey common rules when exchanging messages.
In general, the computers at the two ends of every link on the internet
(Hosts, Nodes and Routers) will ‘run’ the IP (e.g. run a program that
embodies the IP rules so as to forward packets towards their destination).
The IP is sometimes described as a ‘link level’ protocol because it
manages packets as they travel along specific links. IP embodies the rules
for how a node will receive and send packets of data. Remember too that
each IP program running on every node in the network will need to make
its own routing decisions about which link to forward the packet to – e.g.
which node to send it on to.
The TCP, in contrast, is called an ‘end-to-end’ protocol and the rules
are followed by the source and destination Hosts. TCP’s job is to prepare
data for transmission, breaking it up into ‘segments’ and pass it to IP
for the journey. On arrival TCP running on the destination host will
reassemble the segments to create the original data. But if any part is
missing (a lost or corrupted packet), or other problems occur, then the
TCP program on the destination Host can ask for a retransmission by the
source Host.
5.2.2 IP packet switching

When a packet of data arrives at a node, the node’s principal task is to
read the destination address in the header of the packet and to make a
decision as to how to direct the packet ‘towards’ that destination. Perhaps
it already knows how to reach that address in a single link and it can
dispatch the packet directly on its way. But, if not, then the node will
decide on the basis of a ‘best guess’. That is, try to send it to the node it
can connect to which has the best chance of leading to the destination.
This decision can be made based on the first part of the address which
relates to the network that the destination belongs to. In other words, if
the address is in ‘Australia’ then all it needs to do is sent ‘towards’ that
network. If the node is not in Australia itself, then sending it towards the
backbone probably is the best route to choose.
As a packet switching protocol IP has two core parts that are standardised,
first the standards for the packets of data it will handle and second the
standards for an IP address it will use in routing. We can call these the
protocols ‘data structures’. There is then the third aspect of how routing of
packets at the nodes occurs, which we consider later in the chapter.
Figure 5.2 is a schematic picture of an IP packet, made up of six 32-
bit words = 24 bytes of header data, plus the payload (the data to be
transmitted from a sending Host to a receiving Host).
55
}
Version Header Service

Total length 16 bits word 1
(4 bits) (4 bits) (8 bits)
3-bit 13-bit fragment

Identification 16 bits word 2
flag offset
Time-to-live 8 bit Protocol 8 bit Header checksum 16 bit word 3 header
32-bit Source IP Address word 4
32-bit Destination IP Address word 5
Options word 6
Data (payload) word 7...n
Figure 5.2: IPv4 packet structure.

You do not need to consider the purpose of all fields in the IP packet
shown in Figure 5.2 (we have greyed out some), but we will look at the
following to get a better idea of how data is prepared and packaged and
what IP does.
Version: This tells you if this is an IPv4 or an IPv6 packet. This is
required so the IP program knows how the rest of the data is structured. In
the description here we focus on IPv4 packets.
Header: Number of 32-bit words in the header – in this case of IPv4 it is
six words.
Service: This allows for the network to treat different packets in different
ways, e.g. route them differently or give some priority over others e.g.
real time audio data and video data (e.g. a Skype call) may have higher
priority than bulk data transfers or file downloads.
Total length of packet. Maximum is 63,535 bytes (216 – 1).
Time to live: Maximum number of hops a packet can make (max. 255).
This is decreased by one for every hop and ensures that if a packet gets
lost, or is going in circles in the network, that it will in the end by deleted.
When a packet is deleted, a message is sent back to the Sender to say that
the packet was destroyed. The Sender can then resend it.
Protocol: This designates the ‘higher level’ protocol this packet is
supporting, for example, six for TCP and 17 for UDP (more about this
below).
Header checksum: A sum of all the header words, to ensure that any
corruption in the header is caught.
Addresses: The header of an IP packet includes both the Senders address
as 32 bits and the Recipients address as 32 bits. IP addresses in IPv4 are
32-bit binary numbers (e.g. 4 bytes), so the absolute number of separate
addressees is ‘only’ 232 (that is over four thousand million). It is a lot, but
not enough if every device, person, computer, washing machine and candy
bar or medicine in the world is to have its own unique address (think of
the IoT (see Chapter 14). So, in IPv6, the ‘new’ version of IP, each address
is 128 bits – a allowing huge number of separate addresses – we discuss
IPv4 and IPv6 further below.
56
An IP address (IPv4) is usually written for human use as four numbers

between 0 and 255, representing four 8-bit bytes (discussed in Chapter
9). For example, to find the IP address of the University of London
website http://www.london.ac.uk/ we can use the tracert (= Trace Route)
command on a Windows computer with the result shown below (Figure
5.3). To do this yourself search for ‘Command’ in the windows search box
to find Windows old fashioned ‘Command Prompt’ interface. On a Mac
look in the Applications folder and then Utilities folder and then run
the Terminal program. Then type:
traceroute www.london.ac.uk
You can also find tracert apps for smartphones that can do the same task
but you will need to download them first.
Figure 5.3: Screen shot of the Windows tracert program finding the route and
hops to get from my computer to www.london.ac.uk
The screenshot above shows that the International Programmes website has
IP address 151.101.130.216. It also shows that the nice human address we
all use and can remember: http://www.london.ac.uk/ is just a proxy for
a rather more complicated URL address of 3i2pw67lodcps.edge.platform.
sh. The tracert screen shows all the hops across the network from node to
node that a packet makes to get from me at home using my ISP connection
to this website – the first line is the hop from my computer to my router
(dsldevice) at home (192.168.1.1). Note that all the later nodes shown have
a .net domain – until we arrive at the final .ac.uk Host.
Activity 5.1: Using Tracert and investigating IP addresses

Try to reproduce the data shown in the screen shot above, the Tracert program makes use
of the Time-to-live feature of the IP packet (see above). Traceroute sends to the destination
a set of three packets with the Time-to-live set as 1, then three packets with it set to 2 and
then three with 3 and so on. The Tracert program will get three messages back from the
first, second, third…nth nodes on the journey saying they deleted the packet because the
Time-to-live was exhausted and reporting the particular node’s IP address. These addresses
can then be looked up using reverse DNS to find associated URLs. In each case the program
will work out how long (in ms, milliseconds = 1/1000 of a second) it took each packet to
reach that node and for the ‘deleted’ report message to get back. Given that 3 packets are
sent out with each Time-to-live value, you see 3 times reported per node.
Of course, I made the above query and edited this section of the guide in London in
September 2022. You will be somewhere else in the world, on a different date, and have
a different and perhaps longer route to the website. It might also be that the IP address
has changed since then – we will come back to this problem and its solution when we
discuss DNS below. Even if I repeat the task within a week, I may well find that my
57
packets take a different route because each node along the way will make its own best
decision as to the right route depending on the conditions on the day and anyway the
network and the services on it are always being changed and adjusted.
5.2.3 What about IP routing?

So far, we have discussed packets and addresses and looked at how
messages pass between nodes, but not much about the details of the
switching or routing that takes place at each and every node.
To recap, the way the internet works is that packets are generated by the
Sender and then passed to their local network node (router or switch). It
is the responsibility of that node to decide where to send the packet next
on its first (or next) hop – e.g. to route it hence these switches are called
routers.
Nodes read the address in the header of the packet (the network address
and Host ID). Perhaps it already knows how to reach that address – that is
has seen it before and kept details in a local cache or routing table. If so,
it can dispatch the packet directly on its way. But if not, then it will pass it
to a ‘best guess’ node it can connect to on the basis of the first part of the
address which relates to the network that the destination belongs to. In
other words, if the address is in ‘Australia’ (.au) then all it needs to do is
sent the packet towards that network.
Figure 5.4 is a basic routing algorithm adapted from Wikipedia. It
refers to the Node having a ‘routing table’. This is a list of destinations
and the right route (e.g. link) to use. This table can be updated by an
administrator (human being), or it can be updated using what are called
Dynamic Routing protocols to gather information from other routers on
the network. In this course we do not need to dig any deeper into Dynamic
Routing. For us it is enough to assume that the node has a table of routes
for some addresses, but if not then it uses the default.
Given an IP address that is for Network N and host D

if ( N matches a directly connected network address
)
Deliver packet to D over that network link;
else if ( The routing table contains a route for N
)
Send packet to the next-hop address listed in the
routing table;
else if ( There exists a default route )
Send packet to the default route;
else
Send a forwarding error message to the originator;
Figure 5.4: A basic routing algorithm for an internet node.

Source: Adapted from: https://en.wikipedia.org/wiki/IP_forwarding
58
5.2.4 IPv4 and IPv6

The best source for information is Wikipedia at: https://en.wikipedia.org/
wiki/IPv4.
See also the Linus Sebastian’s Techquickie video ‘IPv4 vs IPv6 as fast as
possible’ (2014) https://www.youtube.com/watch?v=aor29pGhlFE.
At various places in this chapter and previous ones we have mentioned
that IP version 4 is the most widely used, but that there is a new version IP
version 6 also in use. The primary difference we have noted so far is that
IP addresses in IPv4 are 32 bits long and in IPv6 they are 128 bits So very
many more addresses are possible.
IPv4 is a stable protocol that still operates according to the specification
set out in September 1981 and indeed is very recognisably in the original
specification of 1974. The stability of these ideas over 40 years once again
indicates how insightful the original developers of TCP were. Of course,
it has had a number of improvements and extensions over the period – as
listed on the Wikipedia page in the ‘Development’ section.
IPv6 dates to around 1998 and it was the response of the internet
Engineering Taskforce (IET) to foreseen address space problems. As well
as more bits for addresses it also readjusts the way that IP addresses are
assigned and makes it more efficient for a router to decode them. IPv6 also
has more security features and makes it easier for a new node to share
information (e.g. routing information) with other nodes around it.
The address space limitation and the need for more bits in addresses, is
usually linked to the coming of the internet of Things (IoT) (see Chapter
14) and the need to give IP addresses to many many many devices and
objects. Anything electronic from a toaster to a torch or a camera will
want an IP address. Something like an aeroplane jet engine might have
thousands of components each of which wants to send and receive
messages, report data, receive program updates etc.
Still and despite the very good reasons to develop a new IP, extend the
address space and offer better security, Wikipedia reports that 99% of
internet traffic in 2014 used IPv4. On the other side they also report that
in 2016 14.1% of Google traffic came via IPv6 – which suggest that things
are moving on. Perhaps there is a lesson or two here about infrastructures
in general.
Lesson 1: If it works for users then they have few reasons to change or
move on.
Lesson 2: The ability of IPv4 itself to adapt over time (a sign of its clever
design in at the outset) has meant that most significant problems found
along the way have been addressed by add-ons and work arounds.
Lesson 3: Infrastructures draw on their installed base and while that base
is not IPv6-ready, there will be little movement. As and when the installed
base is IPv6 ready, the shift may come very fast – an example of a network
effect (see Chapter 2).
Activity 5.2: IPv4 and IPv6

Consult Wikipedia and other online resources about the two IP versions – IPv4 and IPv6.
Compare both and discus the advantages and limitations of each one.
59
5.2.5 The TCP

The best online source for detail of TCP is Wikipedia: https://en.wikipedia.
org/wiki/Transmission_Control_Protocol.
So far, we have considered the IP – the rules that allow packets to move
across the internet from node to node and thus allow messages to be sent
and more importantly received. But you will recall that the internet is not
100% reliable. As we mention in Chapter 3, the internet is a ‘best effort’
network and in Chapter 4, we noted that we should not expect that all
packets for a single message or user will necessarily all travel along the same
route, arrive in the order sent, or even arrive at all. As we have seen when
discussing tracert, the ‘time-to-live’ feature of the IP header is designed
explicitly to delete packets that get lost or are taking too many hops.
This unreliability may or may not be much of a problem. If you have
connected a thermometer to the internet at the South Pole via satellite and
asked it to send a temperature reading every five seconds, then perhaps
you would not worry too much if the odd reading gets lost. Similarly, if you
are sending a video image as part of a live conference call over the internet
(e.g. using Zoom, Microsoft Teams, or Skype), you can live with a few lost
packets of data. The image will flicker or freeze, but the conversation goes
on. This is better than a totally frozen image while you are waiting for a
missing bit of video data to arrive. But, if we send the Word file of this
subject guide to our editor over the internet, we would be very disappointed
if it did not arrive whole and as exactly the same string of bits that we sent.
So, in the case of my Word file, indeed in most but not all cases, we need
to do something about the reliability. We may also have the problem of
the sender sending packets too fast for the receiver to process them – the
problem of flow control. These are issues that TCP addresses and as such
it is layered on top of IP. That is, it assumes IP to be working well enough
and it then adds some more useful functionality ‘on top’ to make the
network even more usable – or we might say to increase the utility of the
digital infrastructure. Later in this Chapter we will look briefly at the main
alternative to TCP – a protocol known as UDP (User Datagram Protocol)
and which adds a different mix of facilities to the raw internet and IP.
Some basics on TCP: TCP is a Host-to-Host protocol and it is also
sometimes called a Transport protocol (for the Transport layer). The
internet is a ‘network of networks’ seen at the IP layer, but once we get ‘up’
to TCP we are considering the internet as only one network and assuming
that it mostly works to get packets from Sender to Receiver – to transport
them. As shown in Figure 5.1, TCP is about the relationship between a
sender Host and a destination Host. The network, e.g. IP and below, does
not care about or do anything much with TCP. The switching nodes and
routers are (mostly) unaware of any higher protocols being used. Indeed,
as mentioned above, some IP packets carry TCP data and some carry UDP
data. Many applications use a mixture of both UDP and TCP, for example
the VOIP service Skype uses both protocols. So, what does TCP do to help
get data across a network in a more reliable way?
In very simple terms TCP does the following – using transmission of my
Word file as an example:
1. TCP on the sending Host accepts as input a data stream (e.g. my
Word file), splits it into chunks, adds a TCP header to each chunk.
The results, chunks of data plus a header, is what is called a TCP
segment. (See Figure 5.5.)
2. The segment is passed to IP (together with the destination IP address).
The IP puts the segment in one (or more) IP packets and sends it on its
60
way across the network. It is generally best if the segment goes in one
IP packet – e.g. we can set TCP to make small enough segments to fit in
one IP packet.
3. The IP packet header will contain the Protocol code that says that this
packet contains data related to a TCP connection. So, when it arrives at
the destination Host, the data (TCP segment) will be handed from the
IP layer to TCP.
4. TCP layer at the receiving Host will acknowledge the arrival of
the data by sending back to the sending Host a TCP segment (of course
using IP) that is an acknowledgement of data received up to the byte
number given in the Sequence Number field. This acknowledgement
message can also set the ‘window size,’ e.g. the number of bytes the
receiver is prepared to handle at one time. In this way the receiving
Host can indicate that ‘I am busy and need less data sent at a time,’ or ‘I
am happy to receive more data’, or even ‘Stop sending me data’.
5. The TCP layer at the receiving host will reassemble the original message
(in our case a Word file) out of the contents of the many segments it
receives, each with a sequence number and then direct the data to the
program designated in the ‘destination port’ field.
TCP segment
Source port (16 bits) Destination port (16 bits) Word 1
Sequence number (32 bits) Word 2
Acknowledgement number (32 bits) Word 3
Data
Reserve Word 4
offset Window size (16 bits)
dwr
psh
ack
ece
urg
syn
fin
rst
ns
(3 bits)
(4 bits)
Checksum (16 bits) Urgent pointer (16 bits) Word 5
Options (32 bits) Word 6
Figure 5.5: TCP segment header.

Source: Based on Wikipedia: https://en.wikipedia.org/wiki/Transmission_Control_
Protocol
The key parts of this header are as follows (noting the set of 1-bit flags in
Word 4 which convey a lot of information about what the segment contains
– these are listed below):
Source and destination ports: A TCP segment includes a source port
and a destination port field – which says which program they originate from
at the source and which program they should be routed to at the destination
– remembering that an IP link to the internet may be sending and receiving
data from a number of different programs at the same time (e.g. email,
WWW, chat, etc.), all mixed together, the fundamental strength of packet
switching and it ability to multitiplex on a data link.
Sequence number: This number indicates the correct place of this
segment in the sequence of segments that make up the message. If the SYN
flag in word 4 is set (e.g. = 1) then this is the initial sequence number to set
up the connection. Actual data will follow in the next segments using this
sequence number incremented by 1 each time.
Acknowledgement number: If the ACK flag is set then this value is
the next sequence number that the receiver is expecting (e.g. the number
61
of the next byte of data expected). This is sent back to the sender and
acknowledges receipt of all bytes of data up to this one.
Data offset: This is the size of the header in 32-bit words. Minimum 5,
maximum 15, e.g. it tells the TCP receiver how many words to skip to find
the payload data.
Option flags: The flags in Word 3 have many uses. So far, we have
spoken of SYN and ACK and below we mention URG. The others we do
not consider.
Window size: This specifies the number of bytes of data that the receiver
is willing to receive (e.g. for flow control). It might be as low as 0 which
means stop sending.
Checksum: A checksum for error checking in the received segment.
Urgent pointer: If URG flag is set this is an offset from the sequence
number indicating the last urgent data byte.
Options (if data offset >5): This option allows the TCP segment
to carry more information in the header for purposes such as setting
maximum segment size, agreeing acknowledgement rules. Remember, the
internet was set up to be capable of being developed and extended and
this is an example of how new features might be tried out in TCP without
preventing the previous rules still being perfectly workable. Achieving
this kind of flexibility is one way to allow backward compatibility
e.g. making experiments, developments and improvements in a way that
still allows older systems or ways of working to continue to operate. It is
fundamental to all infrastructures and in particular digital infrastructures,
that they evolve and develop over time and this makes being able to
support backward compatibility very important.
Activity 5.3: TCP functions and processes

In your own words, define the functions of TCP, giving details to the technical processes
used by the protocol. Particularly explain how TCP works as a transport protocol (end-
to-end), controlling what is sent and what is received, keeping the integrity of the
information sent by the internet.
5.2.6 TCP in operation

The details of the operation of a TCP connection at the level of the
programmed behaviours of the Sender and Receiver are quite complex and
we do not expect students to have a detailed knowledge of this. However,
some basic properties of a TCP connection do need to be understood.
TCP makes the internet more reliable because of the mechanisms we have
seen here:
1. TCP segments are placed in IP packets and IP does its best to deliver
them.
2. TCP data arrives in segments with sequence numbers. The sequence
numbers allow the receiver to reassemble a message, even if the IP
packets arrive muddled up.
3. If the sender does not receive an acknowledgement message in a
reasonable time (e.g. if the IP packets get lost on the way there,
or ACK messages are lost on the way back), then the sender will
retransmit the segments.
4. If the Receiver is receiving too much data to handle, it can send back
an acknowledgement asking for a smaller window, e.g. for the sender
62
to send less data (even no more data) before stopping and waiting for
an acknowledgement.
The Wikipedia page on TCP https://en.wikipedia.org/wiki/Transmission_
Control_Protocol also lists a number of vulnerabilities of TCP – features
that may make it possible to attack or interrupt TCP traffic. This is not a
reason not to use TCP, but perhaps to use it with other security features
and security protocols that then sit on top of TCP. For example, when
you use a secure link in a web browser, e.g. the URL address shows as
HTTPS:// and says ‘Secure’ in your browser, another layer of protocol on
top of TCP is working to encrypt the web page data before it is divided
into segments by TCP.
5.2.7 The UDP protocol

See UDP in Wikipedia at https://en.wikipedia.org/wiki/User_Datagram_
Protocol.
As we have mentioned above, the TCP sits alongside another protocol UDP
(User Datagram Protocol) in the usual internet layered model.
This is a much simpler protocol than TCP as each UDP segment (called a
datagram) is independent of any other. The header, as with TCP, includes
source and destination ports to identify the sending and receiving software
in the two hosts, a checksum and a length in bytes (see Figure 5.6).
Indeed, the source port and the checksum are optional, so UDP is the very
simplest way to send data.
There are no acknowledgements, no set up procedures for a connection,
just some data delivered to a destination and all that is known is the
destination port that it came to – even the source is optional.
UDP Header
Source port (16 bits) optional Destination port (16 bits) Word 1
Length (16 bits) Checksum (16 bits) optional Word 2
Figure 5.6: UDP segment header.

Source: based on Wikipedia, https://en.wikipedia.org/wiki/User_Datagram_Protocol
As you may expect this is a very efficient way to send small or urgent data
if you do not need flow control or reliable transmission. For example, this
might be the case for streaming video across the internet. If it is ‘real time’
as in a Skype or FaceTime call, using UDP may help to ensure a faster link
and a better user experience even if the video or audio quality is at times
a bit rough. Much voice and video traffic on the internet uses UDP for this
reason.
UDP also suits simple small message exchanges. For example, the DNS
system discussed in Chapter 4 uses UDP rather than TCP. Efficiency is
probably the main reason, given the billions of DNS calls that are made
every minute of the day, but which just consist of a simple URL sent and an
IP address returned – in both cases around 10–20 bytes of data in total.
Once again, this choice and the ability to make your own trade-off
between efficiency, timeliness and reliability, are examples of the
generative capability that has been nurtured in the internet world. This is
a network that offers a bundle of features but lets the user to decide how
or when to use them.
63

In this chapter you have been introduced to the three main protocols that
underpin the functioning of the internet. They show you how the internet
works as a network-of-networks that provides basic facilities to get data
from point A to point B (IP) and to do so reliably and controllably (TCP),
but with an option for fast though less reliable transmission (UDP). This
chapter introduced you to IP packets, as well as TCP segments and UDP
Datagrams. Each of these is a data structure that is used to contain the
data to be transmitted. UDP Datagrams and TCP segments will always go
inside 1 or more IP packets before they are transmitted across the network.

Having studied this chapter, and completed the Essential readings and
• explain the character and purpose of a network protocol in general
• give a logical description of the specific tasks that IP, TCP and UDP
support as part of the internet suite of protocols
• understand the main developments that arise as the internet moves
from IP version 4 (IPv4) to IP version 6 (IPv6).

1. Explain the distinction between a link-level and an end-to-end protocol.
In your answer explain what the internet needs protocols of both types
and give an example of each type.
2. Why do both TCP and IP each in their own ways use a system
of sending messages with data in one direction and sending
acknowledgements in the other. In each case explain, how this is done
and what it is intended to achieve.
3. The Wikipedia entry for UDP says that ‘UDP is suitable for purposes
where error checking and correction are either not necessary or
are performed in the application’ and ‘Time-sensitive applications
often use UDP because dropping packets is preferable to waiting for
delayed packets, which may not be an option in a real-time system’.
Explain what these two sentences mean, giving real-world examples to
illustrate your ideas.
64
Chapter 6: Using a layered model of infrastructure: ‘above and below’
Chapter 6: Using a layered model of

infrastructure: ‘above and below’
Chapter outline
This chapter presents the layered model of infrastructure, sometimes also
known as the ‘hourglass architecture’.
This is a general way to understand how pieces of technology are
combined together, each doing a specific task, so as to allow the creation
(or emergence) of the digital infrastructures that people use. The chapter
draws upon the work of Jonathan Zittrain, presented in the book The
future of the internet – and how to stop it (2009). In this book, Zittrain
focuses first on the internet and the way layers are combined to create
the internet and its related services. Zittrain’s argument leads on to the
concept of the internet as a source of innovation, which he presents using
the concept of generativity.
We try to provide as simple as possible an introduction to these ideas,
summarising Zittrain’s ideas to facilitate your understanding at an
introductory level. When you have grasped the basic ideas set out in this
chapter you will be ready to move on to the fuller arguments presented in
Zittrain’s book.
6.1 Introduction
This chapter introduces a general layered model of digital infrastructures.
A ‘model’ in the way we use the word here is a way of framing our view
of the world and helping us make better sense of it. As we all know,
the world is a very complex and messy place. Humans feel better when
they put some order on the world, through classifying and framing the
phenomena they observe and developing rational accounts to explain facts
and events. This kind of modelling is the basis of most science, but also
social sciences such as management, economics, or sociology. If and when
we can test a model against the world, we can then generalise it further to
become a ‘theory’. In this chapter, however, we stay at the level of a model
to help us better to cope with the complexity of our subject of digital
infrastructures.
In this chapter, we present first the layered model of an infrastructure.
Then, using this model, we present the idea of generativity.
We start by presenting Zittrain’s perspective on the layered model
(his hourglass architecture), as he applies it to the internet, but here
we want to be a bit more general and talk about all manner of digital
infrastructures. We then turn to consider how the flexibility of the many
parts enables innovation. In particular, we connect the discussion of the
layered model with the concept of generativity, exploring Zittrain’s
framework that identifies the characteristics of a digital infrastructure
that make it more prone to foster innovation. At this point, we resume the
discussion on the concept of innovation, which is a fundamental aspect
of the discussion in later chapters (as introduced in Chapter 2).
You already have seen a simple introduction to the layered perspective at
the end of Chapter 2, when we show how different pieces of technology
– the internet, the WWW, cloud computing, mobile technologies and
social media – are related to each other and in Chapter 5. In this chapter,
65
we present this formally as a conceptual perspective. In the following

chapters, 7, 8 and 9, we add other important concepts and frameworks to
develop our model further. These four chapters (6, 7, 8 and 9) together
provide you with a theoretical and conceptual framework you use to
understand in later chapters the link between digital technologies,
innovation and change.

• Describe the layered model of an infrastructure using the internet as
an example.
• Present Zittrain’s concept of generative pattern and related
characteristics of an infrastructure.
• Explore how the generative characteristics of the infrastructures are
related to the concept of innovation.

• outline the concepts of hourglass architecture and generative patterns
• use the skills you have acquired to apply the hourglass architecture to
the internet
• explain the relevance of the hourglass architecture to fostering
innovation
• explaining key ideas on how the layered model is relevant for
understanding digital infrastructures.

You may find a copy of this book online through Google Search or Google
Scholar.

Application Programming Interface: https://en.wikipedia.org/wiki/
Application_programming_interface

The chapter presents a conceptual model of digital infrastructures,
illustrated by the hourglass model and generative framework proposed by
Zittrain (2009). The ideas introduced hare are then used in the arguments
presented in the following chapters. Thus, the purpose of this chapter is to
present the layered model here so as to help you understand each of the
other digital infrastructures we discuss in the guide.
6.2 Chapter content

6.2.1 The hourglass architecture
Zittrain describes the internet as made up of by many layers and sub-
layers performing specific functions, which interact with each other
across interfaces. The success of the internet comes from both the
independence of these layers one from another and from the careful
definition and standardization of the interfaces between the layers. This
layering enables the flexible choice and development of the components
within each layer and for layers to be improved and amended with their
66
own timescales as long as the interface between the layers is not changed.
In Chapter 2 we considered a very simple example of an infrastructure
interface when we spoke of the standardisation of the electrical plug and
socket and the voltage of electricity supplied. This allows you to plug
in all manner of different devices to the same electricity infrastructure
in your home. You can move from a fan heater to a vacuum cleaner to
an air conditioner to a radio and the infrastructure does not need to
know or change as long as you use the same plug and work on the same
voltage. Interfaces and standards are then used where different parts of an
infrastructure ‘plug’ together.
The simplest layered model for any digital infrastructure has three layers
as introduced in Chapter 5 for the case of the internet. At the bottom, we
have the physical layer, in the middle the protocol layer and at the top the
application layer. To do useful work we need all three layers to be present
and working (Figure 6.1). The principle of layers is used here at a quite
simple level – just three layers. Later on, we will use similar models, but
with more detail and more layers. For now, we keep it simple. Laudon and
Laudon (2021) show a layered version of the TCP/IP as we discussed in
Chapter 5 above.
We now consider each layer of our simple model on its own, from the top,
then the bottom, then the middle.
Applications
Protocol
Physical
Figure 6.1: Basic three-layered model for a digital infrastructure.
The physical layer is all the cables and wireless transmitters, the
servers and the devices which allow communications to take place.
These are the hardware aspects of an infrastructure and for any kind
of digital communication to occur (other than mindreading) you need
some equipment so that bit-strings are transmitted and received. The
physical layer does not know of or interfere in the sort of communication
taking place or care about the content and it may include many kinds
of transmission technologies suitable for transmission in different
circumstances.
Activity 6.1 Innovation in the physical layer

Research the physical layer protocol called ‘IP over Avian Carriers’ by a search on
Wikipedia. Just how innovative do you feel this protocol is? What exact circumstances is
it useful for?
The physical layer is of course important, it is the foundation of any digital

infrastructure, but it is there in order to allow other things to happen (e.g.
in the layers above and the world in general). In other words, a super-
67
fast high-capacity fibre optic cable between London and New York is of
no use to anybody on its own. It needs users with uses in mind. This is
neatly expressed in the writing of the US author Henry Thoreau who in
1854 wrote the following about technology in general and information
infrastructure in particular:
Our inventions are wont to be pretty toys, which distract our
attention from serious things. They are but improved means
to an unimproved end… We are in great haste to construct a
magnetic telegraph from Maine to Texas; but Maine and Texas,
it may be, have nothing important to communicate. (Henry
David Thoreau, in Walden, 1854, Chapter 1.)
If we turn to Thoreau’s question of ‘something important to communicate’,

we need to consider our model’s top level, the application layer.
The application layer is where the user does useful things that involve
communications of data. For example, to send an email to the University
of London, update a Facebook page, transmit a large file of research data
from a government department in Tokyo to a laboratory in Finland, or
speak to a friend in Cape Town from you home in Buenos Aires using
Skype – perhaps even to report the weather conditions in Maine to an
airport in Texas as a plane departs.
Each of these tasks (and thousands and thousands more you could think
of), require some specific software to achieve the specific task, but they
will all use the physical layer to get their data (bit-strings) transmitted to
somebody at the other end. So, our model, in this case of the Skype call,
can be elaborated as occurring in two matched layers as shown below,
with the network seen as a cloud or black box that links the two locations.
Girlfriend in Buenos Aires - using Skype Boyfriend in Cape Town - using Skype
Applications Applications
Protocol Protocol
Network
Physical Physical
Figure 6.2: Example of application layer using lower layers.

Note that the solid lines here represent the data moving from a
microphone all the way to a loud-speaker. The dashed lines represent the
way that the application at one end (the Skype program) relates to the
Skype program at the other end – that is the Skype program sends data
that makes sense to the other Skype program (voice data). Similarly, the
other layers relate to their peers (their equivalents) with data that they
understand – so its application to application, protocol to protocol and
physical to physical.
This leaves us the protocol layer – the jam in the sandwich – lying
between application and physical. What is it supposed to do?
From our perspective of digital infrastructures, it is the most important
and interesting layer of all. We do not much care what people do using
a digital infrastructure (in this course at least – it is different in IS1060
and other courses) and we do not really want to be electronics or
communications engineers either – focused on the physical layer. What we
68
want to know is how the infrastructure works as an infrastructure and

to understand this we need to investigate the protocol layer.
We have seen that there are all manners of applications that users need and
they are all different. So too there are many kinds of physical equipment
and devices that can support transmission of data, on cables, fibre optics,
on smartcards, by wireless, by satellite or even by bar codes. (Yes, really,
quite a lot of data is transmitted using a bar code – look on your groceries
or medicines).
But and it is a big but, each application will want to use a different mix
of physical layer resources. It is then the job of the protocol layer to allow
multiple different applications to access their desired physical resources
for communications. For instance, your Skype program needs to access
the physical layer to send your digitised voice to Cape Town. And looking
from the other end, to allow various physical resources to access specific
applications (for example when voice data arrives in Cape Town it needs to
‘find’ the Skype program so the friend can hear a voice). So, in the diagram
above the vertical solid (red) lines thus go ‘through’ the protocol layer and
the protocol layer needs to direct data to the right place in the layers above
and below.
However, the protocol layer has a bigger role to play than just linking
up top with bottom. It also takes responsibility for managing the passing of
the data from one place to another and from one application program to
the right counterpart (Skype data to Skype, email data to email, file data to
a file etc.). In general terms you might think of ‘place’ as the right building
(e.g. Host) and application as the right ‘department’ or ‘room’.
In the case of the internet (always our first exemplar for a digital
infrastructure) and as introduced in Chapter 5, the protocol layer is based
around two key protocols. First the internet protocol (IP) and second the
transmission control protocol (TCP). The two go together as the core of the
internet so people often write about the internet as based on TCP/IP – and
in our simple model we can also bundle them together as belonging to the
protocol layer. However, as we have already said in Chapter 5, they are
doing different jobs and one is link-level, while the other is end-to-end.
To recap, TCP/IP defines the way that data is packaged up and transmitted
across the internet – from place to place (IP) and from program to program
(TCP) – thus TCP sits as a layer ‘above’ IP and TCP has an interface to the
application layer program.
Experience of the internet teaches us that the protocol layer needs to be
particularly stable and standardised for the other layers to work well and
the infrastructure to be able to serve many users and many uses. As you
learned in Chapter 5, today the internet works mostly with a version of
TCP/IP that is over 20 years old. This kind of stability allows all kinds
of users of the internet and builders of software and hardware, to make
decisions and pursue investments fairly sure that not much will change.
Or that when and if it does there will be backward compatibility retained.
Thus, today the internet works with two IP – version 4 (widely used) and
version 6 (slowly coming into use).
6.2.2 Generative architecture

A key outcome of a layered architecture with well specified interfaces is
that it reduces the need to comprehend complexity. Developers
and innovators can work on individual layers or protocols within the
layers without having to understand in detail the complexity of the whole
structure and without demanding that the whole structure changes because
69
of their development. With the communication between the parts (the

interfaces) well defined, professionals can specialise in developing a
particular protocol or element in a layer.
For instance, a programmer can develop a new app for a mobile phone
just knowing the communication interface of that device. For developing
a new app for an Android device, the programmer just needs to know the
API (Application Programming Interface) of Android. The API allows the
creation of apps which have access to a particular operating system and
its datasets including how to access defined functions and procedures
for the interchange of data. This architecture speeds up the capacity for
developing new solutions that use the internet. In a similar way, a country
can invest in improving the backbone of its broadband infrastructure
without impacting the apps and the content providers in the network.
As Zittrain argues, the layered architecture fosters innovation and
constant improvement, as groups responsible for various parts of a big
infrastructure can work on a particular element without affecting any
other. This flexibility comes from the limited interdependency of the parts.
There are of course standards to be followed (we discuss standards further
in Chapter 8), but their purpose is as much to allow innovation as to
constrain it.
In addition, the layered architecture allows new users/devices or
protocols to easily join an infrastructure, fostering continuous
growth and increasing the influence of a given infrastructure. For instance,
the more people want to use the internet, the more ISPs (internet Service
Providers) will join the network to provide services. The more people
have smartphones with apps, the more developers can join the network
providing their new apps or adding new programming languages which
are compatible with the other standards. The layered architecture allows
this sort of growth, in which it is easy to provide additional services and
support additional users – to scale up.
The hourglass architecture
In Zittrain’s book, he describes the layered model as applied to the
internet, but it adds a further element to it when he describes it as an
‘hourglass architecture’. His version of the layered model has a broad
application layer with lots of different things available for use and it
has a broad physical layer, with lots of different ways to transmit bits
around the world, but the protocol layer is smaller, less varied. Indeed,
as suggested above, in the case of the internet the core of the protocol
layer is found in just a few protocols – TCP, IP and UDP. Because so many
people, organisations and applications use this small and stable set of
protocols, they ‘insulate’ applications from physical data transmission. This
ensures that we can all share a common set of connections and that those
connections can develop as fast or as slow as technology allows without
interfering with the applications we use. The image of an hourglass is one
of a broad top and a broad bottom, but a small centre, a small waist.
Table 6.1 shows the variety of elements in each layer as Zittrain set them
out, noting that he gives the IP layer a central position in the internet
architecture – we have chosen to bundle it with TCP and UDP. You can
also see that in this version our basic three-layer model as presented above
is broken down into finer gradations of layers which contain alternative
possible protocols, standards and technologies to use. All the elements
both above (applications) and below (physical) are dependent on the
internet protocol – it is thus unavoidable and universal (note for example
that TCP does have alternative protocols such as UDP).
70
Layer Elements
Applications Email, WWW, Skype, Instagram, apps
SMPT, HTTP, HTTPS, RTP
Protocol TCP, UDP
Internet protocol (IP)
Physical Ethernet, PPP
CSMA, async, sonet
Copper, fibre, radio, pigeon
Table 6.3: The hourglass architecture.

Table adapted from Zittrain (2009, p.68). Note that we have made a slight
adjustment to include TCP, UDP and IP in the protocol layer.
Activity 6.2: Exploring the hourglass architecture

In the section above, we have presented the hourglass architecture of the internet. In the
book, Zittrain applies the same model to understand the development of the personal
computer (PC). Read Chapter 4 where he discusses the PC case. Using the two examples,
internet and PC computers, think about the core set of common concepts that this
architecture uses.
Zittrain argues that the hourglass architecture demonstrates a form of

organising an infrastructure which is more generative, in the sense of
being more flexible and welcoming of new elements being created on
the top of an existing infrastructure. As defined by Zittrain (2009, p.70):
‘Generativity is a system’s capacity to produce unanticipated change
through unfiltered contributions from broad and varied audiences.’
A generative architecture for a digital infrastructure is thus very different
from a closed and hierarchical architecture. In a hierarchical structure,
developments are proposed and managed through the hierarchy and any
development needs to follow the guidance and the coherence defined
by the hierarchy. For instance, Microsoft develops and sells proprietary
software which is conceived, designed and built by their employees. As
a user you can buy it and you can change the configuration a bit, but
you cannot change the programme inside. Even making suggestions for
changes you want or need will be difficult and probably unproductive.
With a generative architecture for an infrastructure and assuming a certain
level of openness of information, there is freedom for individual people
and organisations to create new things that can draw on lower layers and
interact with higher layers, without asking anybody’s permission. The only
limitation is that the new elements should follow the basic set of rules of
the system as a whole including using standard interfaces.
In other words, by design and conception, the layered architecture
model produces a generative environment which fosters creativity and
innovation, allowing people to adapt the infrastructure to their needs.
Zittrain (2009, p.71) further argues that the generative potential of a
71
particular digital infrastructure depends on five key factors, as presented

below:
• How extensively a system or technology leverages a set of possible
tasks.
• How well it can be adapted to a range of tasks.
• How easily new contributors can master it.
• How accessible it is to those ready and able to build on it.
• How transferable any changes are to others.
We can summarise these required factors as: leverage, adaptability, easy of
mastery, accessibility and transferability. The next short paragraphs explain
each of these features, summarised from Zittrain, 2009 (Chapter 4).
Leverage: is the capacity for doing useful things. The argument is that the
more things a system is capable of doing, the more chances it has to bring
forward change (innovation). Think about the potential of a clean piece of
paper. How many different things can be written or drawn in this paper?
Leverage means the technology allows itself to be used in different ways.
Adaptability: is the capacity to be adapted easily to a range of different
uses beyond the uses envisaged when developed. The author emphasises
that the adaptability depends on how much a tool or system can go beyond
what has been planned originally. Electricity, for instance, is a technology
that has great level of adaptability.
Easy of mastery: is the capacity a system has to be understood by
many people who either want to use it or want to adapt it to other goals.
Aeroplanes, for instance, have low levels of mastery, anyone who wants to
be a pilot must have extensive training. Modern social media software, in
contrast, is mastered by billions of people across the world with almost no
formal instruction.
Accessibility: is the capacity of being accessible to people and to have
information related to the system available to those who want to learn
and explore the possibilities of the system. Costs, taxes and regulations
can create barriers to accessibility. Digital technologies and services
become more accessible as costs of technology have gone down (think of
smartphones).
Transferability: is the capacity of allowing changes and new insights
or uses to be transferred easily to others. Changes made by users and
specialists can be transferred to others even when they do not have much
knowledge about the system itself. Software improvements, for instance,
are very transferable and software updates are a daily occurrence on
internet-attached devices.
These features are individually desirable, but more importantly, they
support each other. The more an infrastructure has of these features, the
greater its chances of being generative and so drawing in users, fostering
more innovation and creating change. In contrast to non-generative or less-
generative systems, which depend mainly on the creativity and innovation
of their original makers, generative systems free up the bottlenecks of
change by inviting more people and different people to participate in
the process of creation of new solutions and finding new uses for the
infrastructure. (This is also seen on the growth of open-source software
movement.)
The key output of generative systems/infrastructures is innovation, and we
investigate this idea and develop the argument further in the next chapters.
72
We have said before, the internet is an important example of a digital

infrastructure and it is also an important example of a generative system
that has demonstrated its superiority in this respect over other forms of
infrastructure. Before the internet, there were many attempts to build
proprietary and commercial data networks. But they were controlled
by a few people and organisations who sought to define how these
networks would be used and developed. For instance, a bank could
build a proprietary network to facilitate the transaction with key clients
– their own private digital infrastructure. But the bank would define the
functionalities and the nodes of this network and would tightly control the
way this network would evolve.
Digital infrastructures challenge this logic. They suggest the need for
openness. where anyone can connect and disconnect and use them for
any purpose. There is no need for a management hierarchy defining what
can be on what part of the internet and who can innovate elements of
the internet and so to for other infrastructures discussed too to varying
degrees. Thus, innovation comes easily and missing elements such as an
appropriate levels of security for banking data, or reliability for video
streaming, can be added to the layered architecture with new or improved
protocols.
Activity 6.3: Exploring the generative architecture of the internet

Drawing upon the definition of generative systems, explain how the internet has the
characteristics of leverage, adaptability, easy of mastery, accessibility and transferability,
giving your own examples to illustrate your points. You need to read Zittrain’s Chapter 4
to be able to answer this question.
Activity 6.4: Learning the generative power of the mobile infrastructures

In the section above, we have briefly discussed the generative characteristics of mobile
technologies. Go online and research how many apps the different mobile operating
systems (such as iOS, Android, Windows etc.) can provide. Look for the most popular apps
in each of the operating systems. What are your conclusions from these numbers?
6.2.3 The layered model applied to the study of digital

infrastructures
In this chapter we have focused on applying a theoretical perspective
or model to the study of the digital infrastructures. From the discussion
in the previous paragraphs, it should be clear that the internet has
a sort of architecture which facilitates innovation. Drawing upon
Zittrain’s arguments, we can extend the concept of layers to other digital
infrastructures. Cloud computing, mobile technologies and social media
each have a high generative potential and they have become themselves
important digital infrastructures built on top of the internet as elaborations
in the application layer.
Cloud computing, for example, works on top of the internet. Anyone with
access to the internet can access cloud-based services as a continuity of
their normal activities. The cloud infrastructure of servers, virtualisation
(a way of better using servers), data storage and sharable software is itself
generative, allowing in turn the development of other useful applications.
Imagine, a start-up company developing a new game can provide service
to millions of users by use of the cloud. More users mean more processing
power and more storage, all paid for when needed as a service. Such a
company would have a tough time to get enough capital to invest in their
own infrastructure to launch a game. The cloud solution not only speeds up
73
the innovation process, but also reduces the risk for the entrepreneur, who
will pay for the software and the servers in the cloud only if and when the
company has sold the product.
Similar generative patterns can be seeing on social media. Again, social
media services are successful because they operated on the top of the
internet. They become themselves digital infrastructures when offer space
for further innovation on the top of their own layer. For instance, Facebook
has made its API public, allowing developers to design apps to work
within Facebook. Zynga has become a massive player in the production
of games through developing apps that run on Facebook, exploiting its
social network effects. When Facebook makes its API public, it becomes a
digital platform, allowing other apps to be developed on the top of its own
infrastructure (see more details on digital platforms in Chapter 17).
Mobile technologies repeat the same pattern. First, as discussed in
Chapter 2, the key factor for the mobile phones to become so pervasive
is the capacity of using them for accessing the internet and data. There
is thus a symbiotic interaction between the internet and the mobile
devices, in which one leverages the position of the other. In addition,
mobile technologies have become themselves digital infrastructures
with generative patterns. Let’s again think about the APIs of the iPhone
and Android devices. As the APIs are available, developers around the
world have created millions of apps to mobile devices, innovating in the
way services are provided in all areas, from games and entertainment to
education and healthcare.
These three examples show the relevance of the layered model, hourglass
architecture and the concept of generativity for a better understanding
of the nature of digital infrastructures as channels for innovation. In
later chapters we come back to these concepts, to discuss in more depth
these key digital infrastructures. Let’s now discuss two other examples
of solutions created on the top of these digital infrastructures, to better
embed our knowledge on their generative power of the hourglass
architecture of digital infrastructures.
The first example is about delivery services, which can be contracted
by individuals or companies, as per their demand. For instance, you are
at home on Saturday evening with friends, buy your pizzas online and
contract Uber Eats or Deliveroo or your local delivery service to deliver the
pizzas. The app for delivery operates on the top of the mobile operating
system, using the API of Android and iOS, for instance. The app also uses
the power of cloud computing, as all exchange of information between
the users and service providers happens in the cloud. The delivery
companies have incentives to use cloud services, as they have irregular
flow of data in their systems, depending on the demand at any given time.
The cloud services (such as Amazon Web Services) allow the company
to expand their use of servers when necessary. This has happened during
the Covid-19 lockdowns, when many individuals and families used the
delivery services to buy food from restaurants which were operating only
with takeaways and delivery. Indeed, many restaurants have changed
their operating model during the Covid-19 pandemic, moving from
serving food inside to providing takeaway food. These restaurants have
quickly innovated their business models joining platforms for delivery. The
innovation from the restaurant’s perspective was easy and fast, because
the digital infrastructures and the digital platforms for delivery were
already there, ready for absorbing as many new restaurants as required.
The second example is also about the response to the Covid-19 pandemic.
Governments around the world have created, through their health
74
systems, apps for citizens to be informed about the current Covid-19

isolation rules, or for individuals to register whether they were sick (post-
testing). In the United Kingdom, for instance, the NHS (National Health
Service) in England created an app to track and trace people who had
contact with others found to have Covid-19. Using Bluetooth, the app
was able to identify whether a person would have been close enough
to an infected person. This innovation of apps for tracking and tracing
during the Covid-19 pandemic used the generative power of mobile
infrastructures (including Bluetooth), the APIs provided by the mobile
platforms (such as Android and iOS) and the cloud infrastructure to collect
and process the information. All these digital layers were there before
the Covid-19 pandemic. The work of governments was to understand the
need of communicating with citizens and tracing the spread of the diseases
and to ’cultivate’ the digital infrastructures and platforms to create the
necessary solutions. The innovation was enabled by the generative
potential of layers of digital infrastructures and platforms.

This chapter presented the layered model of infrastructure and the
hourglass architecture and discussed the concept of innovation. The
chapter shows how carefully structured but loosely coupled pieces of
software and hardware allow the internet to thrive, becoming the most
important digital infrastructure. The chapter emphasises that the hourglass
architecture, organised in layers, has generative characteristics that
foster innovation. Following Zittrain (2009), the chapter explains the five
characteristics which make a digital infrastructure generative: leverage,
adaptability, easy of mastery, accessibility and transferability.

• outline the concepts of hourglass architecture and generative patterns
• explaining key ideas on how the layered model is relevant for
understanding digital infrastructures
• use the skills you have acquired to apply the hourglass architecture to
the internet and other infrastructures
• explain the relevance of the layered model and hourglass architecture
to fostering innovation.

1. What is the particular significance of the hourglass architecture for i)
innovations and ii) for scale for an infrastructure??
2. Define the five features of generative digital infrastructures including
your own examples to illustrate your points.
3. Explain what is meant by generativity. Illustrate your answer with two
examples of a generative digital infrastructure other than the internet
75
Notes
76
Chapter 7: Core ideas for evolving infrastructures I: installed base and cultivation
Chapter 7: Core ideas for evolving

infrastructures I: installed base and
cultivation
Chapter outline
This chapter explores theoretical concepts to help understand how digital
infrastructures evolve. The chapter makes a strong claim that it is difficult
if not impossible to ‘design’ digital infrastructures. Rather, that (successful)
digital infrastructures evolve over time and in unexpected ways, generally
starting their development by drawing upon other infrastructures.
Through time, layers are refined, adapted and perhaps added moving
towards a digital infrastructure that is generative – that is self-powering.
To explore these ideas, the chapter uses the two concepts of installed
base and cultivation, drawing upon examples from the internet and
other digital technologies.
7.1 Introduction
In Chapter 6, we discussed the characteristics of digital infrastructures,
using the layered model and hourglass architecture. This perspective is
useful to understand the current state of art of these digital infrastructures.
However, we have not yet discussed how powerful digital infrastructures
are created. This chapter and the next focus on understanding how digital
infrastructures are developed, providing some theoretical concepts to be
used later when we analyse other infrastructures.
In this chapter, we present two key concepts: installed base and
cultivation. First, we need to understand that key digital infrastructures
evolve through time. Despite our best efforts to design an infrastructure,
the fact is that things are going to change in directions we have never
thought. Thus, we always need to leave room to accommodate the
new developments while making the best use of an installed base.
Second, as digital infrastructures are not designed from scratch, we
need to understand their process of development, which may be more
about accommodating change in face of the innovation, a process which
has been called cultivation by scholars. Lastly, we introduce the idea
that institutions influence and are influenced by digital infrastructures,
pointing out that digital infrastructures affect human behaviour, a topic to
be discussed further in Chapter 9.

• Explore how digital infrastructures evolve.
• Explain the limitations of design when facing installed bases of
previous technology and new needs.
• Define the concepts of installed base and cultivation.
• Explore the implications of these concepts for our understanding on
how digital infrastructures evolve.
77

• explain the difference between design and cultivation when analysing
digital infrastructures
• describe the impact of installed bases in the development of digital
infrastructures
• discuss the capacity to apply the concept of installed bases in particular
contexts
• discuss the capacity to apply the concept of cultivation to better
understand digital infrastructures.

You should read one of the articles below. It is not necessary to read both,
as the key concepts are repeated. However, you can find some examples
which are different in each article.
Ciborra, C.U. and O. Hanseth ‘From tool to Gestell: Agendas for managing the
information infrastructure’, Information Technology & People 11(4) 1998,
pp.305–27. Available in the Online Library.
Hanseth, O. ‘From systems and tools to networks and infrastructures – from
design to cultivation. Towards a theory of ICT solutions and its design
methodology implications’ (2002) Available online: https://www.igi-global.
com/gateway/chapter/44242

Ciborra, C.U. ‘The platform organization: recombining strategies, structures,
and surprises’, Organization Science 7(2) 1996, pp.103–118.
Ciborra, C.U. and G.F. Lanzara ‘Formative contexts and information technology:
Understanding the dynamics of innovation in organizations’, Accounting,
Management and Information Technologies 4(2) 1994, pp.61–86.
Hanseth, O. ‘Inscribing behaviour in information infrastructure standards’,
Accounting, Management and Information Technologies 7(4) 1997, pp.183–
211. Available in the Online Library.
Minitel: https://en.wikipedia.org/wiki/Minitel

Ciborra, C. The labyrinths of information – challenging the wisdom of systems.
(Oxford: Oxford University Press, 2004) [ISBN 9780199275267].
Particularly Chapter 4 (Gestell). Most of the content in this chapter is based
on the concepts developed by Ciborra. Examples and additional concepts
have been also used

The chapter explores key concepts for understanding digital
infrastructures. We start arguing that digital infrastructures evolve
in directions nobody could plan or control, especially when we are
talking about the key infrastructures discussed in this subject guide. In
this way, we challenge the idea of a simple focus on design for digital
infrastructures. In order to explain the way digital infrastructures
evolves, we present the concept of installed base. Through examples
and theoretical arguments, we show that digital infrastructures evolve
on the top of installed base layers and each new layer needs to take
into consideration the limitations and potentials of the previous ones.
In the sequence, we propose to use the metaphor of cultivation as an
alternative view of the concept of design. The chapter finishes with a
small introduction to the idea of that institutions affect the way digital
78 infrastructures evolve, a topic which is further developed later.
7.2 Chapter content

An interesting question to be asked about key digital infrastructures is
about how they have been designed to become what they are today.
The question comes from a limited comparison between a corporate
infrastructure and infrastructures which are used across corporations
and countries, involving millions or even billons of people. Within
an organisation, management may try to have control of its digital
infrastructure, by design. Even in this case, it would be difficult to have
absolute control of corporate infrastructure in a centralised manner,
considering the many layers which support such a system. However, the
logic that applies to an organisation does not apply for massive digital
infrastructures, which in fact emerge from the interactions of a huge
number of social actors, from computer scientists, programmers and
professionals, to organisations, governments and lots of different users.
7.2.1 Digital infrastructures evolve

Digital infrastructures – as we see in this subject guide – are not in general
designed from scratch by a single set of people. They emerge from and
develop by drawing on people’s needs and their creativity, from the
interaction of social actors in accordance with various contexts of use.
Infrastructures evolve from one level of technology to another through
periods of time – think back to the 100s of years of evolution of our
railway networks or electrical power grids, each of them slowly taking on
new technologies and pensioning off older ones. There is indeed some
conceptualisation of any infrastructure in the beginning, but this is a seed
for further development and any digital infrastructure, we will argue, is
going to evolve in accordance with different needs and alignments that
emerge and as new technologies are introduced. Think how often the
social media infrastructures that you use add new features?
For instance, the internet – which has evolved to become the network for
billions of people nowadays – relied on basic telephone infrastructure to
start operating. At that stage, the physical layer supporting the internet
was just capable of sending low volumes of data such as instant messages
and emails. Even photos were too slow to be downloaded through such
physical infrastructure. From that base, the internet has evolved to the
current state in which a substantial amount of the internet traffic is of
videos and TV watched on mobile devices, which rely on very high-speed
wireless broadband connections. The original internet was imagined as
able to transport much less data at much slower speeds than today. But
time, new needs and technical innovations have allowed it to change.
The same pattern can be seen in the applications which run on the top of
the internet. Initially, the commercial internet was mainly to communicate
content online, especially text, with a few photos and images. Gradually
these applications have evolved too. Search engines and the ‘google’
business model has been created, facilitating finding relevant content. The
cost of data transmission and storage has gone down, making it possible
to have access to large images and streaming videos. This allows services
like YouTube to emerge. The technology has more generally changed to
allow interaction and conversation (the emergence of Web 2.0, which we
will discuss in more detail in Chapter 13), thus enabling the creation and
use of digital social networks. E-commerce services like Amazon and eBay
have become viable as a mass of buyers and users are online and relatively
safe online payment systems have been created.
79
This short history of innovation, growth and change shows that the digital
infrastructures we know it today has not been designed. Rather, it has
emerged from the original system, which already was built relying on
relied on other earlier infrastructures (such as the telephone). And there
is more to come we can be certain. Our digital infrastructures will keep
on evolving in unforeseen ways. Thus, the concept of design in a holistic
sense is not that useful for us as we try to understand the emergence of
digital infrastructure. We need a different framework, as discussed below.
Activity 7.1: Back to the history of the internet

Research the initial concept of the internet as a network of networks using the link below
(see the topic, The Initial internetting Concepts). Find out how the open architecture
networking allows a scalable digital infrastructure to emerge from different digital
networks which can evolve in their own way. In which ways does this open architecture
support the argument that the internet has not been designed from scratch?
https://www.internetsociety.org/internet/history-internet/brief-history-internet/
7.2.2 Installed base

The first relevant concept here is the idea of the installed base.
Defined simply, the installed base is all the actual current elements of an
infrastructure at the current time – literally what is installed and in use.
From their origins, as we have said in earlier chapters, infrastructures
are built by using existing resources and then adapting them over time.
This is another way of saying that digital infrastructures are not designed
from scratch, they are created on a much more modest scale, using earlier
technology and practices that are available. Remember that the internet
was created out of a small number of existing networks. It is the nature
of digital infrastructures that it takes time for them to emerge in their
early manifestation and as time goes on, the uses made of them and the
requirements will change, challenging wherever plan there was at the
beginning. To use a non-digital example, the road infrastructure in many
countries has had to adapt to heavier trucks (e.g. stronger bridges) and
faster driving speeds (e.g. widening and straightening roads).
In the beginning, nobody can be sure a particular infrastructure is going
to become relevant and sustainable in the long term. The problem then
often arises that the sort of development which is necessary in later stages
(e.g. heavy trucks on the roads) may not be compatible with the original
installed base (e.g. basic low strength bridges). Most of the times it is not
viable to eliminate all the installed base and to start again from scratch –
although there may be examples. In other words, the installed base defines
a path dependency for the way the whole digital infrastructure is going
to evolve across decades to come. Any new innovations come gradually,
trying to integrate with previous practices, standards protocols etc. (more
on standards in Chapter 8). One idea relevant for digital infrastructures is
to add new layers that can incorporate the new needs and compensate for
limitations of the previous layers. Still, this could mean losing some of the
benefits of being interoperable with the other parts of the infrastructure.
An example of adding a layer, or more strictly adapting a protocol, is when
encryption security is ‘added ‘on-top’ of TCP. See for example TCPCRYP at
https://en.wikipedia.org/wiki/Tcpcrypt
Another challenge comes from the number of elements in the layers of a
digital infrastructure. The more elements, the more difficult it becomes
to negotiate changes in any level. Go back to the hourglass architecture
discussed in Chapter 6. You can see that each element of each layer has its
80
own standards, with its own applications, which are supported by different
stakeholders with different interests in that particular piece of technology.
The larger the number of stakeholders, holding different interests, the
more complex the negotiation for change, especially if it is necessary to
obtain an agreement among social actors who hold contradictory interests.
For this reason, the ideal model for large digital infrastructures is the one
based on maximum flexibility on how parts are to be developed and used,
fostering the interoperability through the use of standards which regulate
the forms of interactions and communication among the parts. The ‘tight
waist’ of the hourglass, with its limited and stable protocols also allows for
a good insulation of innovation and change above from innovation and
change below.
Ciborra (2004) particularly discusses these aspects of digital
infrastructures, proposing alternative approaches for managing the
phenomenon of complexity. Ciborra proposes to accept the characteristics
of digital infrastructures and adopt strategies such as constructing
gateways and adapters (as the internet pioneers did with their IMP
device; see Chapter 3). The information flow is then driven through these
gateways and adapters. In this fashion, each part of a large network would
be allowed to evolve in its on terms and gateways would connect these
heterogeneous parts in order to form a bigger network which is functional,
in spite of the differences of the parts. Again, internet features such as the
backbone and the DNS serve this kind of strategy.
Through gateways, it is possible to connect and guarantee the
interoperability of all parts, even those which are very peripheral to
the system as a whole. This process makes possible the inclusion of less
relevant or well-resourced parts in a network, thus providing for those
small groups which may require something very specific for their needs. It
also increases the flexibility of the whole network, as new modules can be
added to the main system without affecting directly all the other modules.
Again, we see here an opportunity for innovation. Indeed, as the user
base grows and diversifies, we can expect to observe more of the network
effects: with users’ needs and their desired use influencing in the way the
digital infrastructure evolves.
This approach helps to shape a better interaction with the installed base.
First, it recognises that there is a diverse installed base of a myriad of
technologies and that new developments need to both build on of this
and reshape it. It recognises that it is almost impossible to substitute
an installed base, either for reasons of costs or for political reasons (the
different perspectives of stakeholders). In other words, in recognising
that there is an installed base which is not going to change easily, we then
accept that that each part of a digital infrastructure will evolve to some
degree in its own way or at its own speed and the focus of professionals
and users should largely be in creating gateways and adapters for the
infrastructure to better work together as a whole. Easier said than done.
Ciborra argues, unsurprisingly, that the internet is a good example of
how a digital infrastructure grows through involving many different
stakeholders and pieces of technology. He suggests that the internet has
evolved out of the perspectives of two strong groups. On the one hand
the military people, looking for a resilient communication network which
could survive a massive attack from an enemy and on the other hand the
academic community, which would look for a technology which would
favour decentralisation and open exchange of data. TCP/IP appears
as a good solution to bring together the property of resilience and a
decentralised network, although both key stakeholders would continue
81
have very different ideas on the desired outcome and uses of this design.
Indeed, TCP/IP has allowed different networks to come together as a
unique digital infrastructure, by providing the gateways through which
intercommunication is possible and thereby gained the momentum
necessary to allow the quick growth of the user base and the innovation and
evolution of the installed base.
7.2.3 Cultivation substitutes design

Conceptualising the situation described above, Ciborra says that it
is necessary to change the perspective on how to manage digital
infrastructures. Technical and managerial perspectives would expect a
system to be designed first. However, analysis of the economic aspects of
digital infrastructures recommends a different approach: the cultivation of
the installed base. That is, start with what you have and we work to make
incremental improvements, draw in new users and provide valuable and
sharable services.
This concept of cultivation comes from the perspective that control of
digital infrastructures is very difficult, particularly when talking about
large infrastructures, which are the focus of this subject guide. The idea of
cultivation recognises that each part of a large infrastructure can evolve
in its own direction. In this context, formal design or policing of digital
infrastructures may not be successful. Instead, accepting the challenges
imposed by the installed base and its evolution, the idea of cultivation
suggests that we need to take into consideration how the elements and
layers of a particular digital infrastructure impose their own limitations on
the other elements and layers. Any strategy of ‘going against the installed
base’ is unlikely to succeed in a very complex situation, as it is often
impossible to redesign the whole infrastructure from scratch.
Activity 7.2 Cultivating IPv6

Consider the case of IPv4 and IPv6. The installed base has IPv4 and is not very keen to
shift to IPv6 – what would you suggest as ways to ‘cultivate’ the internet so that it does
get to IPv6 in the end?
Cultivation means that neither humans nor technology are isolated

elements in such complex networks. It also recognises that the best strategic
approach for digital infrastructures is to understand what has been defined
or allowed for in their initial design (the range of possible features and
configurations). Ciborra and other associated authors refer to this range of
possibilities as inscriptions. These inscriptions foster a particular range
of behaviours (patterns of use or programmes of action) and outcomes, in
such a way that people cannot do much outside of this range of possibilities
when they are using a particular piece of technology. In practice, this
concept of inscription means that technology may impose on users a
particular range of behaviours, which is not only present in the particular
layer or element of the infrastructure but also in the other parts of the
network which are related to this particular layer and element.
From this perspective, in accordance with Ciborra, the stability of a network
depends on the negotiation between the many social actors and pieces of
technology, to get a better alignment of interests between all stakeholders.
This process of alignment is possible through a sort of reinterpretation of
one’s interests in relation to the others. Ciborra uses the term translation
to express this situation in which social actors reinterpret their interests thus
keeping the necessary alignment to keep the digital infrastructure operating
as a whole, integrated system. After a strong alignment of interests is
82
established, it becomes more difficult to change any part of this aligned

infrastructure. However, this alignment is also fundamental for a digital
infrastructure to grow. Thus, translation of different interests into some
features, to be inscribed within the digital technology, is a fundamental
step for any digital infrastructure to become relevant to and be used by big
numbers of users.
Activity 7.3 Inscriptions, translation and alignment of interests

Try to think of some features of a social media infrastructure that major stakeholders
may agree upon (align their interests) and want inscribed into the infrastructure – think
of four groups negotiating on this: users such as you, advertisers, social media site
owners and governments and regulators. Then identify some things you may not all
agree upon.
Developing the idea of cultivation, Ciborra argues that a digital

infrastructure should be “planted in fertile soil” (Ciborra, 2004, p.69).
Certainly, the installed base on the top of which new infrastructures
operates need to allow grow of the infrastructure. The author uses the
example of the WWW, which has been designed to operate on the top of
the existing internet – e.g. the internet is ‘fertile soil’. As we have already
discussed earlier, the network is capable of transmitting messages in an
ordered way. However, it does not facilitate finding specific information
online. The WWW defines the structure of addresses of the content and
allows the finding of content, through the URLs and hyperlinks. The
interaction between the internet and the WWW is such that many people
do not differentiate between them, referring to both as ‘the internet’.
Another interesting historical example of this concern about having fertile
soil is given by Minitel. In a nutshell, Minitel was an earlier infrastructure
developed in France, like the internet, allowing people to consult to
electronic data (telephone directories and stock prices, for instance) and
defined transactions, such as purchases of train tickets. As we know, these
activities later have been developed on the internet. In the Minitel network,
some pieces of information would be for free and others would be paid
by credit cards or in the telephone bill. Companies could provide online
services via the Minitel terminal, for a fee. In spite of the similarities,
Minitel had a very different architecture, when compared to the internet.
It was a closed and proprietary network, designed and owned by the
national telecom company, which has kept the control of its development.
Its soil was not so fertile. In 2012, France Telecom closed the service, as
the internet had become the pervasive digital network of our times – but
Minitel has been in decline for many years before this final ending.
Activity 7.4: The survival of the fittest

In the previous paragraph we have discussed the idea of fertile soil, to better understand
the concept of cultivation. The Minitel case shows that not all very interesting
infrastructures survive in the long term. In this exercise, you are going to explore the roots
of social media as we know it today. Go to https://en.wikipedia.org/wiki/Usenet (and
beyond) and research the meaning of USENET.
In which way is this technology similar to and different from the current use of internet
forums and social media tools such as Facebook? This is an interesting exercise for you to
understand how technologies are built on the top of other infrastructures and to explore
as well how a particular technology has influenced the development of others, such as
the social media infrastructure and interfaces.
83
7.2.4 Beyond the technical aspects of technology

Ciborra calls attention to a subtler phenomenon when discussing the
emergence of digital infrastructure: the cognitive understanding and
institutional arrangements of a society are going to influence and be
influenced by the emergence of a digital infrastructure. Once interpreted
and used in a particular way, digital infrastructures influence the way we
frame and solve problems, the way we design organisations and forms of
working. It also influences the way we perceive some behaviours and uses
of technology as being legitimate and proper in defined contexts. This
perspective challenges the view that technology is a rational, technical
phenomenon, proposing we need to understand how digital infrastructures
are also institutional arrangements, defined by social interaction between
stakeholders.
Again, the internet provides good examples to help you understand this
theoretical perspective. For instance, we all expect that companies of
relevant size have a presence on the internet, that flight companies sell
tickets online and that companies communicate regularly by email. These
expectations frame the way companies and individuals behave. For instance,
if most of my friends and people in my social circle use social networks, this
creates a pressure for me to use social media as well. For some, it would
even appear weird that one does not use social media in some social groups.
Another example: anyone looking for buying a computer these days would
think first to compare prices on the internet, either in individual vendors or
in comparison websites. The idea of going from store to store researching
prices would be seen as a waste of time.
An interesting aspect of this cognitive and institutional view of digital
infrastructures is that sometimes the new form of doing is not even as
efficient as alternatives. However, the shared interpretation of how things
should be done using digital technology is going to influence behaviour
anyway. An interesting example of an institutionalised behaviour is the
habit of sending text messages through mobiles. Who would imagine that
people would prefer to text long messages, instead of calling a person on
the very same device and sorting out the conversation in a few minutes?
One could argue that this choice is also fostered by the difference in costs,
which favours messaging. However, many subscription services now offer
unlimited calls and texts at affordable prices, and yet people keep writing
messages. Asynchronous communication by message is more acceptable
than being interrupted by a phone call. This is a new social agreement, in
comparison with the previous option (calling).
This institutional perspective is going to be further discussed in Chapter 9.
Activity 7.5: More details on conceptualising digital infrastructure

Go back to your notes related to ctivity 2.1 (Digital infrastructure: key underlying
concepts). Add concepts from the chapter to your list of key concepts.

This chapter challenged the idea that one can design a massive digital
infrastructure. Using theoretical arguments and examples, the chapter
showed the complexity behind the way digital infrastructures evolve.
The concept of installed base and cultivation was used. Examples from
the internet supported the argument. The chapter ended with a short
introduction to the argument that institutions influence and are influenced
by the evolving digital infrastructures, a topic to be discussed later in this
subject guide.
84

Having studied this chapter, and completed the Essential reading and
• explain the difference between design and cultivation when analysing
digital infrastructures
• describe the impact of installed bases in the development of digital
infrastructures
• discuss the capacity to apply the concept of installed bases in particular
contexts
• discuss the capacity to apply the concept of cultivation to better
understand digital infrastructures.

1. Using the example of the internet and Minitel, explain the argument
that digital infrastructures are difficult to design such that they will
have long-term success.
2. Based on the chapter and recommended readings, explain how the
installed base is related to the phenomenon of path dependency.
3. In a situation in which design is not possible, explain how the concept
of cultivation may be useful to explain how digital infrastructures
evolve.
85
Notes
86
Chapter 8: Core ideas for evolving infrastructures II: standards and network externalities
Chapter 8: Core ideas for evolving

infrastructures II: standards and network
externalities
Chapter outline
This chapter keeps the focus on core ideas for allowing infrastructures to
evolve. The focus is on the role of standards, both formal and informal.
Digital infrastructures become key technologies that can scale across the
world and support multiple uses based on their use of clear standards
for exchanging data and information and sharing access to services. How
standards are established and used is fundamental to understanding how
digital infrastructures evolve. In this chapter we see that standards both
enable and constrain the way a digital infrastructure is used and evolves.
The chapter then goes on to explain the idea of network externalities, which
is related to the size of networks and the related challenge of lock-in.
8.1 Introduction
In Chapter 7, we argued that digital infrastructures are difficult or even
impossible to design from scratch. We also argued that the way to develop
or evolve a digital infrastructure is through exploiting the installed base
and by using gateways and adapters add functionality. In this chapter
we say more about how standards for interfaces and protocols perform
this role and allow the different parts (layers, elements in a layer) of a
digital infrastructure to interact and communicate with others. Through
common standards, layers and elements of them can evolve independently
of each other while still keeping the infrastructure working together as a
whole. Enabled by standards, digital infrastructures can grow and gain
momentum to become key players in economies and societies.

• Explain the role of standards in the development of digital
infrastructures, both formal standards and provider defined
application programming interfaces.
• Explain how digital infrastructures are prepared to gain scale and reach.
• Understand the concept of network externalities and its relevance for
• Highlight the risk of lock-in situations and the potential for standards
to ameliorate them.

• understand the type of standards found in digital infrastructures and
their roles
• differentiate open/public and closed standards and APIs
• critically discuss the significance of network externalities for scale and
scope
• identify the risk of lock-in situations and the potential to avoid them.
87

Weitzel, T., D. Beimborn and W. Konig ‘A unified economic model of standard
diffusion: the impact of standardization cost, network effects, and network
topology’, MIS Quarterly 30 2006, pp.489–514.
West, J. ‘The role of standards in the creation and use of information systems.
Standard making: a critical research frontier for information systems’, MIS
Quarterly Special Issue Workshop 2003. Available online: http://www.
joelwest.org/misq-stds/proceedings/148_314-326.pdf
Network externality: https://en.wikipedia.org/wiki/Network_effect

Hanseth, O. and E. Monteiro ‘Inscribing behaviour in information infrastructure
standards’, Accounting, Management and Information Technologies 7(4)
1997, pp.183–211.
Net neutrality: https://en.wikipedia.org/wiki/Net_neutrality
Open standard: https://en.wikipedia.org/wiki/Open_standard
Long tail: https://en.wikipedia.org/wiki/Long_tail
Data portability: https://en.wikipedia.org/wiki/Data_portability

Hanseth, O. and E. Monteiro ‘Inscribing behaviour in information infrastructure
standards’, Accounting, Management and Information Technologies 7(4)
(1997), pp.183–211.
Hanseth, O. ‘Knowledge as infrastructure’ in Avgerou, C., C. Ciborra and F.
Land (eds) The social study of information and communication technology
– innovation, actors and context. (Oxford: Oxford University Press, 2004)
[ISBN 9780199253524] pp.103–118.

The chapter focuses on developing the concept of standards, explaining
how standards allow interaction and communication between the many
layers and elements of a digital infrastructure. The concepts of open and
closed standards are introduced as well as Application Programming
Interfaces (APIs). The example of the internet protocol is used to support
the theoretical arguments. In the sequence, the chapter explains the
phenomenon of network externalities, either with overall benefits or
with overall risks and challenges for all nodes (positive and negative
externalities, respectively). Drawing upon these concepts, the chapter
explores ideas of scale (how much an infrastructure may grow), reach
(how many people and organisations an infrastructure may reach) and
lock-in situations (when an infrastructure locks-in users, making it difficult
for individuals and organisations to use or create alternatives).
8.2 Chapter content

8.2.1 Introduction to standards
All digital infrastructures are based in a myriad of standards, which affect
hardware and software aspects of infrastructures and how users connect
with them. These standards may be defined by a formal agreement
between interested social actors, by law or by the wide acceptance
of an established practice. The most fundamental standard in digital
infrastructures is the one which defines the communication pattern (i.e.
how data is transferred from one point to another in the network – TCP/
IP). A standard also needs some established processes for its improvement
over time. In the case of TCP/IP this is via the Internet Engineering Task
Force (IETF) as discussed in Chapters 4 and 5.
88
Standards are everywhere in our lives. When you buy a kilo of sugar you
have a particular expectation about what is meant by the word kilo. When
you buy a car, you have a particular expectation in terms of how much this
car is going to pollute the environment or how safe it is in a crash. When
you board an airplane, you have a particular expectation about the size of
the bag you can have under your seat or fit in the lockers. Standards thus
facilitate our exchanges and interactions in society, shortcutting the need of
further negotiation at any new exchange or interaction.
Standards represent some aspect of the chosen communication pattern
of a particular digital infrastructure. This is achieved in accordance
with a complex negotiation process in which the social actors involved
(professionals, associations, organisations, governmental and regulatory
bodies etc.) come to a consensus which is going to benefit all. In other words:
all social actors involved in the definition of a standard for a particular
aspect of a digital infrastructure have an interest in the digital infrastructure
working well. This is the starting point of the negotiation. Naturally, as we
have seen in the previous chapters, this conversation may not start from
scratch: the established standards of the installed base may affect and limit
the choices social actors have to propose new or revised standards.
Earlier in this subject guide, we discussed core standards in the
protocol layer, the internet protocol which defines how data packages
are routed across the internet. This standard protocol makes possible
the communication across the network. Any node transmitting and/or
receiving information on the internet benefit from IP as a standardised
way of communicating. You buy a computer, you open the browser and
immediately you have access to the internet and the WWW because your
software adheres to standards such as IP or HTTP, doing it the same way
as billions of other people. This strong adherence to a common standard
means that the network you use is shared with billions of other people,
which increase the value of it to you.
Let’s consider another example, of an area in which standards are not
working so well. Think about your mobile phone charger and how many
you have had in the past few years. Each mobile manufacturer decides
which standard they are going to use for their own charger. They even have
different chargers for different devices. If you need to charge your mobile,
you need to have a charger which is specific for your device (voltage, plug
etc.). It would be easier for most of us if all small chargeable devices could
be charged with a standardised charger. We are certainly not there yet. In
this case, the industry players have not reached an agreement – indeed
it may not be in their interests to do so and users like you and I have not
exercised enough pressure to get full standardisation, which could benefit
us all. Note that in this case there is a wider environmental benefit that may
arise from standardisation in reducing the number of chargers produced
and the number thrown away.
8.2.2 Open and closed standards

A fundamental aspect of technical standards found in digital infrastructures
is whether they are open or closed. An open standard is one which is
publicly available and has proprieties which allow different social actors
to decide how to use the standard. Open standards are generally freely
available and well documented on which grounds they are more likely to
be adopted. A closed standard either imposes a rule that limits access
to it (as in proprietary systems), or just allows some to know or to use the
standard, in accordance with defined criteria (such as payment of fees,
licensing of patents etc.) Technology companies have many standards
89
that they develop and use in their products or services, but which they
carefully control and limit access to. On other occasions they may make
a standard relatively open and freely share it on the grounds that this
helps everybody. Even open standards have diverse levels of openness. For
instance, a standard embodied in open-source software allows the code
that expresses the standard to be seen, used and amended. TCP/IP is an
open standard, available to all to use, which is owned and managed by a
collective body (the IETF).
The APIs of private companies who offer online services and platforms
(e.g. Google, Facebook, Amazon, Dropbox etc.) are an important kind of
closed standard. User companies and individuals have access to the API
and can develop apps to run on these platforms and use their services.
However, the platform owner has the ability to define the rules of use of
the API and thus the platform. When announcing the launch of the App
Store, Steve Jobs, the founder of Apple, set quite tight rules for creating
apps on the platform, with some content and functionality being excluded.
In this way, the Apple platform retained the right of curating the content
which is available in their virtual space and can take a cut of revenues
generated by these apps.
8.2.3 Standards enable and constrain action

From the previous discussion we can see that standards are enablers
of easy communication and exchange, particularly in infrastructures
that work on a large scale and include many players or stakeholders.
The network benefits from all users having a clear standard to follow,
facilitating the interoperability between the various parts of the system
and the whole infrastructure gains flexibility and adaptability. However, at
the same time, the standard constrains the things which can be done in the
network. This is inherent to any piece of technology: in enabling a certain
group of actions, the technology also inhibits or impedes others.
Let’s consider an example from the internet. One of the characteristics of
the internet is that data packages travel in the network with the same level
of priority (Net Neutrality – see Chapter 16). This enables anyone to have
their data travelling on the network with other data, having the same level
of priority. In this way, the internet is a more democratic place, as anyone
can participate in the transmission of content and all may access content
from others. However, this model also creates constraints. Not allowing
priorities means that ISPs are not allowed to give priority to particular
content. ISPs cannot differentiate, for instance, data necessary for running
businesses from data used for gaming and streaming films. Some would
argue that data used by companies and governments should have priority
over data used for gaming and streaming videos. However, if ISPs can
differentiate data, other sorts of arrangements may be made possible. In
giving priorities to companies, regular users may be left behind with a
slower internet, as there may not be incentives to provide faster networks
to households who do not pay as much.
Another risk is that ISPs may sell the right for priority in delivering
data to particular companies. For instance, a news company X could be
interested in paying for having its content online before its key competitor
Y. The competitor Y could be out of the market not because its content
or products are not good, but because internet users would not have
quick access to the content. In addition, it could mean that content from
ordinary people or small and medium companies would never be reached
as they would not be able to pay a fee to the ISPs. For this reason, those
90
who defend the Net Neutrality argue that the benefit of having free and
equal access to information is more important and that the constraint of
not allowing ISPs and other players to get a bigger share of the wealth
generated by the internet must be maintained. In order words, the current
Net Neutrality standard is the one which offers a more beneficial and fairer
distribution of value across the whole network, not discriminating between
big and small users.
The example of Net Neutrality shows that the choice of a particular
standard is not neutral, based only on the best technical solution (see
more in Hanseth and Monteiro, 1997). Any sort of standard for digital
infrastructures – even if just an agreement on how things should work – is
defined in a political environment, in which social actors try to get the best
from their own perspectives, limited by the characteristics of the technology
and their own ability to influence outcomes. In other words, standards
embed defined patterns of use, which have a level of flexibility for sure, but
which also imply limitations. The way a particular standard evolves through
time also depends on the way social actors negotiate and align their wider
interests.
For key digital infrastructures – such as the ones discussed in the following
chapters of this subject guide – the process of standardisation happens at
an international level, involving companies, governments, international
standardisation bodies and professional associations, interest groups and
international institutions. Again, we see that the emergence of standards is
not solely a rational design process; rather, it is a political, social process,
which tries to get a number of social actors to align their interests around
a particular piece of technology or way of doing things. This also shows
that the final decision on a standard is only one possibility among many.
It could be otherwise. However, once a particular standard is chosen and
individuals, companies and institutions start using it, it becomes more
and more difficult to change anything in that standard – it becomes
the constraining installed base. And so we go back to the idea of path
dependency, which we discussed in Chapter 7.
8.2.4 Application Programming Interfaces (API)

There is another kind of standard that we need to discuss. Not the global
standards that underpin the protocol layer of an infrastructure (e.g.
TCP/IP or HTTP), but the more specific kinds of ‘standards’ that allow
specific services to be offered in the applications layer. These are what
are generally known as Application Programming Interfaces (APIs). The
acronym API refers to standardised interfaces between programs or online
services – a software-to-software interface, as opposed to a software-to-
(human) user interface, or a software-to-hardware interface. The term is
used widely in discussion of business computing, in particular business
computing that depends on linking to and using other providers of
services, as is the case in most cloud services (see Chapter 11). APIs are
in many ways similar to formally defined standards, doing a similar job of
offering a well-defined way of accessing a particular cloud service.
As well as cloud service using API, so too do digital platforms such as
Google’s Android, Apple’s iOS, Facebook and Amazon Web Services
(AWS). They use API to allow third parties to develop pieces of software
(apps) which run on the top of these platforms. Without a defined API,
either the platforms would not be viable, they would be able to develop
apps themselves, but not integrate the work of others, or any work for
integration of new apps on the top of the platform would be expensive
91
and time consuming and would require the platform to develop the
communication links with the new add-ons. API also allows third-party
apps to talk to each other too, as far as they use the API of the platform
to exchange information. This adds another layer of potential innovation,
not only for apps to operate ‘on the top’ of a platform but also for apps to
collaborate with each other ‘via’ the platform.
In technical terms, an API is a standardised interface defined (specified) by
the provider of the requested service or services (e.g. a software module
or cloud SaaS provider). The API is published – made known to others
who may wish to use it – and says as it were ‘This is what I can do for
you…if you give me the right data’ (not so dissimilar to what a protocol
such as IP offers). Of course, the requestor of a service needs to know the
specification of the API, so as to access the API in an appropriate way and
with appropriate data. The requestor says, as it were ‘I understand that
you can do these things, … I would now like you to do this particular one
… and here is the data you need’.
Wikipedia describes an API a bit more formally as follows:
‘An application programming interface connects computers or
pieces of software to each other. It is not intended to be used
directly by a person (the end user) other than a computer
programmer who is incorporating it into software. An API is
often made up of different parts which act as tools or services
that are available to the programmer. ‘ https://en.wikipedia.
org/wiki/API
In the last sentence of this definition a useful distinction is made between
a bundle of interfaces that perform related work – the usual use of the
term API – and the narrower focus on one individual service (sometimes
called an ‘end point’). Thus, the Dropbox API is a bundle of specific
interfaces that the cloud storage provider has developed to allow its
clients (other companies/users/apps/programs) to write their own code to
create files, read files, delete fines, make folders and so on, each of these
being individual services within the overall Dropbox API. To get a better
understanding look at the public Dropbox API description (specification)
at https://www.dropbox.com/developers/documentation/http/overview.
Note that you are not expected to know any of the specific detail – just get
the broad idea.
The Wikipedia entry goes on to note that:
‘One purpose of APIs is to hide the internal details of how a
system works, exposing only those parts a programmer will find
useful and keeping them consistent even if the internal details
later change.’
This aspect of APIs is compatible with the layered model and the wider
role of interface standards supporting ideas of cultivation and generativity.
The internet and WWW function by using many protocols and you could
say, APIs (standard interfaces). As we have seen, these help to isolate the
detail of specific functions or services (end-points) and thereby allow the
detail of how things are done to be changed (improved we hope) without
undue consequence elsewhere (again, we hope).
The layered model of a digital infrastructure implies that between layers
there will always be some kind of an interface – for example when TCP
passes an outgoing packet to IP it does so ‘software to software’ and we
might say that it ‘uses the IP module API’. Some of the detail of this API
is expressed in the packet format we discussed in Chapter 5. Similarly, at
92
the receiving host, the IP module will pass the data ‘up’ to the TCP module
using what we might call ‘the TCP API’. Similarly, in the WWW HTTP
provides an API for any web server software to use including the GET(xxx)
command to ask for data from a server and the PUT(xxx) command to
send data to a server. Still, it is probably better to talk about the internet
and WWW in terms of standardised protocols and reserve the term API for
higher level Application layer services.
8.2.5 Application layer view of APIs

APIs we might say are the ‘glue’ that holds the application layer of modern
digital infrastructures together, or perhaps the ‘Lego brick’ is better
metaphor than glue. APIs are created by service providers to establish
a standardised format by which service users can connect to them. APIs
define the ‘rules of the game’, what interactions are possible.
APIs can be:
• Public – widely disseminated so anybody, or almost, can use it to
provide or access a software or service. In such a case the API may
be an open standard and used by many different service providers
working in the same domain.
• Private – only for use within a closed group – perhaps a single
company or specific industrial sector. The detail of API is known only
to members of this group.
But in the world of the business platforms, digital commerce, cloud
services and social media, APIs are often in between public and private –
what Wikipedia calls ‘Partner’. Some people can use them, but there is
some kind of control or registration needed – see example below.
The importance of APIs for this course is not essentially a question of their
specific definition or their technical details, but rather the way they serve
to support a dynamic and evolving digital landscape – how they serve to
allow digital infrastructure to develop over time and exhibit generativity.
We have argued earlier those digital infrastructures need to be accessible
and sharable for all kind of users and uses and need to have the potential
to achieve new features and functions by combining different kinds of
existing digital services and adding new ones. These are fundamental
ideas that allow evolution and exploitation of the installed base. In the
Application layer APIs allow this kind of innovation or generativity to
occur.
If your business offers a service via a well-developed API, one that other
people can understand and use as they develop their own systems, apps
or digital products, then you are ‘open for business’. Other people/
businesses/apps/programs will be able to use your services and perhaps
pay for them.
A simple example: You run a restaurant chain which you present and
advertise on various social media platforms.
Through your social media you want to be able to offer information on
your locations, opening hours and menus. You also want to have facilities
to book a table, order food for delivery, order a taxi to take people to
or from the restaurant, or give directions for driving or taking public
transport. You also need to be able to take payments from credit cards
(payment for takeaway orders, perhaps deposits for reservations). Perhaps
you also want to recruit staff via social media. To do all of this you need a
wide range of digital services to make your social media function well and
drive your business forward.
93
All the services listed can be accessed via the API of other firms offering
relevant services. In London it might be Uber (taxi), Opentable (booking),
Opayo (credit card payments), Deliveroo (delivery), Google maps (driving
and public transport), JobDiva (recruitment). In your own hometown
there will certainly be a different set of possibilities – but the same
principle will apply. To use each of these services you will probably need to
register as a user and usually pay to use the service in some way or other –
though some may be ‘free’ as in you pay with your data.
Even a small business may need to use tens or even hundreds of APIs to
access needed service. The restaurant chain described above will have
more needs for digital services and hence more APIs to use. For example,
in their supply chain of food and other goods needed to run a restaurant,
to run a payroll, to keep accounts, recruit staff etc. There is also a
specialised set of service providers that sell API management services to
big corporations to help them develop API for their services and keep track
of all their connections online.
A new phrase, the API economy, has emerged to register the
importance of APIs for digital businesses and their economic function
in driving revenue for service suppliers and innovators. SaaS providers,
for example Adobe, Microsoft, Salesforce, Dropbox or Zoom, have
products and services that are fundamentally based on their valuable API
that connect them to their customers and drive their revenue. Another
important part of the API economy are intermediaries who ‘match’
transactions between two or more parties. For example, a taxi company
like Uber that matches drivers with passengers via apps on mobile devices
of passengers and of drivers. Indeed, all manner of Digital Business
Platforms (see Chapter 17) are manifestations of the API economy.
Activity 8.1: APIs as interface standards

1. Explain why the Android API is an important standard and how it helped the
operating system to gain momentum and critical mass.
2. Search online for information on the ‘API Economy’. What arguments are put forward
to explain its importance for modern businesses? When searching, look in particular
at the web pages of big management consultancies such as McKinsey and API
management service suppliers such as Mulesoft or Akana.
8.2.6 Network externalities: scale and scope

Why do social actors come to agree on or adopt a particular standard?
Mainly because they want to facilitate a particular communication process
or exchange. One benefit of having an agreed standard is that more
people are likely to adopt that standard. Let’s think about our example of
the kilo. Once this standard has been adopted by companies, consumers
and governments, it is much easier to buy and sell products and to check
whether the negotiation has been fair. There is no question about what is a
kilo. The more people who adopt the same standard, the easier it is for the
whole system to work well.
In economics this is known as a ‘network effect’ or a network externality.
This is the effect that in a network, all members benefit or loose from the
addition of new members. This is called a network externality because it
implies that external actors benefit (or loose) from a particular transaction
even if they have not participated directly in this transaction.
These effects are often positive – a positive network externality – for
example new members on the eBay platform, WeChat or on Facebook
94
increase the value of using the platform for all members. Network effects
can be negative too. More cars on the road equal more congestion and
pollution – a negative network effect. A common negative network
externality for digital infrastructures is seen when cyber criminals choose
to focus on the largest social media networks. The problems of ‘fake
news’ can also be explained in part as a negative network externality.
Network effects, in particular positive network externalities, are often
used to explain why successful digital infrastructures grow very large. The
argument is that size (number of members, nodes) matters and people
recognise that belonging to or using a bigger network delivers more value
to them. Better to be on Facebook (around 3 billion users in late 2021)
than a smaller local social media platform. The argument applies to the
internet too. The internet has become a massive and dominant digital
infrastructure because it has been successful in attracting a large variety of
users and uses – where ‘uses’ indicated scope. The size is explained by the
value that people and services gain from being part of a big network. Note
also that this would not have happened if the internet has not defined
clear standards that could adapt to all kinds of different circumstances and
uses. And the more people, organisations, services and devices that use the
internet, the more benefits the network brings to each connected host.
We also need to be careful not to confuse economies of scale with positive
network externalities. Economies of scale are about the actual costs to
doing the business and there may well be some such economies in a large
digital infrastructure versus a smaller one. Externalities, in contrast, are
about the value (positive or negative) that participants experience. In
economic terms the two can be seen as two parallel concepts: economies
of scale operating on the supply side and externalities on the demand side.
This concept says that the more social actors that use a particular network,
the greater the benefits for all actors in this same network. This is called
a network externality because it implies that external actors benefit
from a particular transaction even if they have not participated directly
in this transaction. This is the case with the internet, as we explain in the
next paragraphs. One version of this is found in Metcalfe’s Law which
states that, ‘the value of a telecommunications network is proportional to
the square of the number of connected users of the system (n2)’ (https://
en.wikipedia.org/wiki/Metcalfe%27s_law).
When the internet was created, its value was very low. It was a
communication system used by a few people around the world in
universities and research labs. At that time, access to the internet was not
open and the system to access the network was not at all user friendly by
today’s standards; not many people had or used email. In 1993, when the
internet was opened to commercial use, commercial providers of email
started offering services. Quite quickly, companies and individuals opened
email accounts. As more companies and individuals joined, others saw
the potential benefit for them and joined too. As more people entered the
network, all members gained benefit.
We also should think about negative impacts, such as cybercrime, breach
of privacy, ‘fake news’ and cyber bullying. These can be called negative
network externalities, as they are damages or potential damages
which increase because more people are in the network. For instance,
if few people use internet banking, criminals would not have the same
incentive to go online trying to trick customers and hack banks and credit
card companies. Some cyberattacks have become iconic examples of the
fragility of the whole system in face of these network externalities. For
instance, in 2011, Sony PlayStation was hacked, affecting 77 million
95
customers around the world. It is quite common to receive emails from

criminals trying to collect our data for further crimes, in addition to the
risk of downloading malware in our computer at any moment.
The adoption of standards and positive network externalities taken together
fosters the growth of digital infrastructures. The standards are a core
technical aspect which allow a particular digital infrastructure to be able to
operate at on large-scale and to reach large numbers of people, organisations
and institutions. Positive network externalities complement this by creating
an impetus for demand to grow as size grows. Taken together these two can
imply that bigger will always be better or more successful. Facebook offers
a good example. For Facebook to grow in supply terms, it is just a matter of
adding more servers which can operate the very same developed system.
Working in a cloud structure, any update to the Facebook systems is done
online, with very low costs. The added servers cope with the increasing
volume of data (each new photo is going to be inside Facebook’s servers).
And, to provide impetus on the demand side, as numbers grow so too does
the value of each user of Facebook, drawing in more and more.
Indeed, the spread of digital goods and services through digital
infrastructures has allowed companies to provide a wider variety of
products, thus increasing the customers’ choice, even though just a small
group of people are interested in each specific product. In the marketing
discipline, this phenomenon is known as the ‘long tail’. In the traditional
economy, of physical goods and services provided in face-to-face interaction,
the cost of providing a variety of products is high. For this reason, mass
standardisation is the strategy for providing lots of people with cheaper
products: as the products are mostly the same, the cost of producing
each product is lower. In the era of digital goods and services, the cost of
customisation is affordable, thus companies can provide products which are
more tailored to the needs of individual customers. Companies can even
attend to those customers who want some very specific service if the cost of
production/customization is low. So, the long tail of customers extends the
reach of digital goods and services because of digital infrastructures.
The internet is the most obvious example of the benefits of
standardisation, but we can apply the same logic for the other technologies
we are discussing in this subject guide: cloud computing, mobile
technology and social media. Dropbox, for instance, is a cloud service.
Anyone can back up and share files using the system, independently of
their location, the operating system of their computers, or the browser
they use to access the internet. The service adopts internet standards to
make access to the company’s servers easy via their API. Mobile phones
are amazingly helpful because they use standards which allow us to
communicate with each other. People using different devices, different
service providers, different operating systems and living in different
countries can all communicate with each other over the same network,
allowing the technology to reach a large number of users. Finally, think
about Facebook. This social network was not the first, but it gained
momentum as its interface has proved to be more popular than others.
However, there was an inflexion point at which so many people where on
Facebook that others wanting to join a social network would also go to
Facebook. People want to be in the place their friends and potential friends
are and the more people join Facebook, the greater the positive network
externalities for all who are members of the network.
96
Activity 8.2: The network externalities of social media

In this section, we have introduced the concept of network externalities: the bigger the
network, the bigger some impacts, whether positive or negative. In this exercise, you
need to critically reflect on your use of Facebook or any other social media channel. What
are the benefits for you which come from the critical mass of users of this service and
the variety of uses it is put to? What are the dis-benefits or risks? Could you think how
Facebook (or other platforms) could help by reducing the negative impacts and increasing
the positive impacts of its service?
8.2.7 Lock-in situation

The example of Facebook is useful for introducing another concept. Before
Facebook became the main social network, other social networks also
offered a similar service. At one point, Facebook was disputing space with
the other social networks, such as Orkut (Google has discontinued the
service after launching Google+, which has also now closed) and Myspace.
There was a moment, however, that Facebook became so relevant in terms
of numbers of users and volume of traffic, that it became more difficult
for other companies to take its place. The more people join Facebook, the
more this service locks-in its users. Where do you want to be? Close to your
friends and family, who are on Facebook. Would you change to another
social network? It might happen. However, you will need to think about
another issue: what is going to happen with the photos and interactions you
have had with your Facebook friends over a number of years (or decades!)?
There are other technologies that can create similar situations. Once
they become very pervasive, it can be too costly for someone to leave the
network. In other words, the switching costs are too high vis-à-vis the
benefits offered by the other technology. Some refer to this as a situation
in which the winner takes all. We prefer to say that the winner takes a
bigger slice of the market share, not necessary the whole market. At least
so far. We can see that Facebook is the most important social media tool
around the globe (but not in China, where WeChat leads the way, or in
Russia where VK is used), but there are many other relevant ones. Android
has got the biggest market share of the mobile devices, but Apple strongly
keeps not only a relevant slice of the market but also a substantial amount
of profits. Big companies are fighting for the cloud computing market,
including Amazon, Google, Apple and Microsoft and IBM, but no player
has come close to excluding the others so far.
In the end, a lock-in situation may make the use of a particular infrastructure
irreversible. This does not mean people could not do otherwise. At any time,
a society may decide to not use a particular infrastructure. However, this
may imply a huge cost (switching costs) which might be impossible to pay
(financially or politically). This is an extreme situation, but we can see some
traces of it in the important digital infrastructures we discuss in this subject
guide. In other cases, the lock-in is temporary, as new generations or social
groups change their perspectives on the use of a particular technology. For
instance, some social groups have migrated their preference from Facebook
to WhatsApp as their main communication channel with closest friends and
family, others to TikTok. From a societal perspective, digital lock-in is not
beneficial. People want to have flexibility to change their digital platforms
and keep the property of their own data over time. Also, if data could be
moved from one platform to another (data portability regulations), there
might be more incentives for companies to be more innovative to keep the
competitive advantage. Without the risk of losing users, platforms may not
have the incentives to innovate as much as otherwise.
97
Activity 8.3: Data portability

The situation of being locked in to a particular digital infrastructure is not ideal. People
and organisations want to be able to migrate their data from one service to another. In
this section we have discussed, for instance, the situation of being locked-in on Facebook.
For this exercise, go online and discover how data portability can prevent the situation of
being locked-in in a particular digital service and not able to migrate your data. Start by
looking at Wikipedia (https://en.wikipedia.org/wiki/Data_portability), but research further
other documents.
Reflect on how society could work to create an environment in which data portability
could be the new standardised practice among platforms (tip: use the concept of APIs
here to discuss the technical architecture for allowing data portability between platforms).
Activity 8.4: More details on conceptualising digital infrastructure

Get back to your notes related to Activities 2.1 and 5.3 (Infrastructure: key underlying
concepts). Using the concepts of this chapter, complement your page with more details.
Keep this page with your notations on key concepts of digital infrastructures for your
future reference. This page will be useful in your exams too.

The chapter explained how open and closed standards allow the interaction
and communication between the many layers and elements of a digital
infrastructure, drawing mainly upon examples of the internet (such as IP).
Relatively large digital infrastructures allow the emergence a phenomenon
called network externalities, which means that the larger the network
(the number of nodes) the greater the positive and negative impacts upon
the users (positive and negative network externalities, respectively). The
chapter also added other concepts, which are key features of large digital
infrastructures: scale, reach and lock-in situations. Through exercises, the
chapter fostered the development of a critical perspective on key presented
concepts.

• explain how standards are necessary to allow a digital infrastructure to
gain critical mass
• conceptualising open and closed standards
• critically discuss working with network externalities
• describe key features of digital infrastructures (scale, reach and lock-in).

1. Explain how standards are important for the development of digital
infrastructures.
2. Drawing upon your experience with Facebook (or other social media
tools), discuss positive and negative externalities of the social network.
3. Explain how the concept of lock-in is useful for understanding the
debate on data portability.
98
Chapter 9: Economics and institutional perspectives on digital infrastructures
Chapter 9: Economics and institutional

perspectives on digital infrastructures
Chapter outline
This chapter introduces some economic and institutional perspectives to
analyse digital infrastructures. It summarises two economic perspectives
that are directly related to digital infrastructures. The first is an economic
perspective called Transaction Cost Theory (sometimes called
Transaction Cost Economics). This helps us see how digital infrastructures
have changed industry structures and helped create important new
marketplaces in the economy. The second perspective explores the specific
character of digital goods and services. This helps us understand
the incentives and challenges faced when producing digital products and
services to be to be distributed through digital infrastructures. In the
third section the chapter introduces the institutional perspective
that can help the study digital infrastructures as they relate to society
and the economy. The section explains how institutions influence
and are influenced by digital infrastructures and how an institutional
interpretation of infrastructure helps explain its wider acceptance or not.
The chapter also introduces briefly an alternative perspective on how to
understand the value of digital infrastructures through the resource-
based view of the firm (RBV or RBVF). This theory helps managers
to think about digital technologies from the perspective of competitive
advantages. The RBV perspective can be easily combined with other
theories based on economics or institutions and it is also compatible
with managerial theories of business strategy which focus on competitive
advantages of firms.
9.1 Introduction
Observing the pervasiveness of digital infrastructures in contemporary
societies, you may ask how we can explain their rapid rise. This chapter
introduces some theoretical perspectives that can help us understand why
digital technologies in general and digital infrastructures in particular
have spread across the world so rapidly. The chapter explores perspectives
from economics and from institutional theory. It is not the objective of this
chapter to give a very detailed account of these theories, rather to give
short summaries of the key ideas. The aim in adopting this approach is
to provide some simple alternative ways to explain the success of digital
infrastructures and give enough background to the ideas so they can be
applied directly in the next chapters. This chapter also connects digital
infrastructures with digital goods considering that digital goods depend
on digital infrastructures to reach end-users and digital goods add value to
digital infrastructures, because they are essential channels for distributing
these goods. For further understanding of each the topics, we recommend
additional readings listed below. The chapter ends with a short
introduction to the resource-based view of the firm (RBV) to emphasise
the strategic value of digital technologies for individual firms as a way to
build competitive advantage.
99

• Introduce the basics of Transaction Cost Theory.
• Introduce the concept of digital goods and their key features.
• Connect these economics perspectives to the topic of digital
infrastructures.
• Explore some institutional perspectives on how social actors influence
and are influence by digital technologies.
• Introduce the resource-based view of the firm.

• explain how to use transaction costs to interpret the diffusion of digital
infrastructures
• apply the basics of the transaction costs perspective to understand how
particular digital infrastructures help suppliers and buyers to better
coordinate the exchange of goods and services
• apply the basics of digital goods perspectives to better understand the
phenomenon of digitalisation of goods
• outline basic ideas of institutional theory and resource-based view of
the firm and apply them to digital technologies.

Transaction costs: https://en.wikipedia.org/wiki/Transaction_cost
Resource-based view (RBV): https://en.wikipedia.org/wiki/Resource-based_
view

Avgerou, C. ‘IT as an institutional actor in developing countries’ in Krishna,
S. and S. Madon (eds) The digital challenge: information technology
in the development context. (London: Ashgate Publishing, 2004)
[ISBN 9780754634454] pp.46–62.
Barney, J. ‘Firm resources and sustained competitive advantage’ Journal of
Management 17(1) 1991, pp.99–120. https://journals.sagepub.com/
doi/10.1177/014920639101700108
Cordella, A. ‘Transaction costs and information systems: does it add up?’,
Journal of Information Technology 21(3) (2006), pp.195–204. https://link.
springer.com/chapter/10.1057/9780230250611_19
Quah, D. ‘Digital goods and the new economy’ (2002). Online resource, at the
London School of Economics and Political Science. http://cep.lse.ac.uk/
pubs/download/dp0563.pdf
Peer-to-peer file sharing: https://en.wikipedia.org/wiki/Peer-to-peer_file_
sharing.

Scott, W.R. Institutions and organizations: ideas, interests and identities.
(London: Thousand Oaks; New Delhi: Sage Publications, 2013) 4th edition
[ISBN 9781452242224].
Silva, L. and J. Backhouse ‘Becoming part of the furniture: the
institutionalisation of information systems’ in Lee, A.S., J. Liebenau and
J.I. DeGross (eds) Information Systems and Qualitative Research. (London:
Chapman & Hall, 1997) [ISBN 9781475754872].
Swanson, E.B. and N.C. Ramiller ‘The organizing vision of information systems
innovation’, Organization Science 8(5) 1997, pp.458–74.
100
Teece, D. and G. Pisano ‘The dynamic capabilities of firms: an introduction’,

Industrial and Corporate Change 3(3) 1994, pp.537–56. https://academic.
oup.com/icc/article/3/3/537/696604?login=true

This chapter presents a short account of three theoretical perspectives
which are useful for understanding digital infrastructures. In the first
section, the Transaction Cost perspective is explained. This perspective
shows at a macroeconomic level why digital infrastructures have spread
so widely. The second section discusses a framework to show the nature
of digital goods, again drawing upon an economics perspective. The
section discusses the need to balance the benefits for innovators and
consumers of digital goods in order to have a greater gain for society as a
whole. The third section introduces an institutional perspective on digital
technologies, to have a glance of how social actors and social structures
influence and are influenced by digital technologies. The last section
introduces the resource-based view of the firm theory (RBV or RBVF),
with emphasis on the strategic perspective on resources for sustainable
competitive advantage.
9.2 Chapter content

9.2.1 The Transaction Cost perspective
The Transaction Cost (TC) perspective is based on the idea that there
are different ways of coordinating the production and negotiation of
goods and services. The key for this understanding is to focus on the way
information flows from one economic agent to another and the way these
economic agents organise the contracts to allow them to keep working
in situations one party is going to trust the other. This perspective is very
useful for the study of digital infrastructures because of its focus on the
flow of information as a mechanism to coordinate production and flow of
goods and services.
Let’s start from a perfect market situation, from the perspective of the
economists. In this market, all agents would have access to all information
available to make better decisions in terms of what to produce and what to
buy, in real time. In this perfect market, the price mechanisms coordinate
production. On the one hand suppliers are going to keep producing goods
up to the moment that they are still making profits for each new item
produced. On the other side, consumers are going to evaluate the price
of the good vis-à-vis the benefit that this good brings (economists use the
concept of utility to define this benefit). The important thing to observe
here is that the price mechanism creates the equilibrium in the market, in
such a way that suppliers and consumers get what they need for a price
that is good for both. In this way, the invisible hand of the market does the
work of optimising the allocation of resources in an economy.
Naturally there is a problem with this view of the way production and
flow of goods and services are coordinated: economic agents (producers/
suppliers and consumers) do not have the same level of information. More
than this: some economic agents will try to get more advantages in the
market by adopting opportunistic behaviours. The traditional example is
to think about the market for used cars. The car dealer may not tell the
whole truth about the vehicle on sale. This opportunistic behaviour, on top
of the overall lack of information on the part of the buyer, creates obstacles
for markets to coordinate the production and flow of goods and services.
101
In face of this challenge, economic agents need to coordinate production

and flow of goods and services in other ways. For example, by increasing
the trust and the flow of information, through the creation of firms
(hierarchies in contrast with markets), companies reduce opportunistic
behaviour, as it is easier to control the transparency of information
inside an organisation. Part of the problem is solved, but we still have
the majority of transactions happening in a market in which information
flow is far from perfect. It is thus important to go in more detail to the
characteristics of this information flow, which is the backbone of the idea
of transaction costs.
In brief, transaction costs are the costs for getting the products and
services transferred from the supplier to the consumer. We can divide
transaction costs broadly into three categories (some authors prefer four
categories though):
Search costs are the costs related to finding the information to allow
the exchange to happen, such as finding the suppliers or the customers,
the best price in the market (the supplier looking for the highest price a
customer may pay and the customer looking for the lowest price a supplier
can provide the good). A simple example is a person looking for a new
mobile phone. Before buying the new device, they are going to look for
information on the sorts of devices, features, problems the devices have
and the prices of different suppliers, to make a better-informed decision.
Naturally this effort varies in accordance with the value and relevance of
a particular good or service. If a product is very cheap, such as a coffee
in the street, one does not care much to research the cheapest price in
the neighbourhood, as the cost of searching would be higher than the
difference in price. Purchasing more expensive or enduring goods and
services would justify more search, more information.
Bargaining costs are the costs required to negotiate the contract
between the seller and the buyer. This cost involves the negotiation itself,
in which each party tries to get a better deal and the writing of the contract
which documents the terms in which the transaction is happening. Some
authors may call this contracting costs. Again, the amount of costs
(in terms of time and resources used) for getting the negotiation and the
contract done depends on the type of product and service. If a person is
buying a house, especially through an estate agent (a third party interested
in the deal between the seller and the buyer), the cost of negotiation is
high, with imperfect information going back and forth from the buyer to the
seller and vice-versa. The actual problems of the house and neighbourhood
are not known by the buyer, who needs to protect their interests through
a better contract, which will depend on having the services of a good legal
representative. Again, for very cheap products, this contract is very quick,
most of the time regulated by laws. In countries in which the rights of
customers have been protected, such as in the United Kingdom, when one
buys a product with a defect the seller is supposed to replace the product
immediately. The purchase receipt is the contract between the parties
to guarantee this right without further negotiation. Thus, the amount of
bargaining costs depends on the sort of product and the sort of regulation
that is valid for that particular market in a given country.
Policing and enforcement costs: are the costs required to be sure
that all parties follow the contract, which may sometimes include the
legal costs involved to take the party to court in order to have the contract
respected. Some authors call this the control and regulation costs.
This cost may happen in any transaction, but they are more likely to
happen in transactions of higher value goods and services, in which the
102
opportunistic interests of parties may be more important. Coming back to

our example of buying a house, the contract may say that the house would
have the kitchen cabinets and appliances installed and the buyer, after
getting the keys, discovers that the kitchen space is empty. Thus, there is a
breach in the original contract and it may require a court case to resolve,
which is a very expensive process. This sort of cost may be less prevalent
when negotiating low-cost goods and services, as neither suppliers nor
buyers are interested in fighting for small amounts of money. Still, this
may happen, for instance, when economic agents take a long time to sort
out differences which are not that significant in terms of money.
Activity 9.1: Applying TC on the internet

Drawing upon the definitions of the three types of transaction costs, pick a particular
e-commerce service – such as Amazon, eBay or Alibaba – and analyse how it supports
the reduction of:
• search
• bargaining
• policing and enforcement costs.
In your answer link different features of these services to the reduction of each of these
costs. It is important to be specific, connecting features with results in terms of reducing
the transaction costs. Look carefully on the online interfaces of these services to find the
right features.
Summarising in one sentence, overall, we can say that digital infrastructures

have reduced transaction costs in many marketplaces – travel, music,
clothing, but also iron ore, oil, wheat or rice. Let’s start with the search
costs. Buyers and sellers use digital infrastructures to find more and better
information in real time. They can compare prices and features of products
and services either going to different websites or using internet services that
specialise in comparing prices. For instance, a person can compare flight
tickets by visiting the websites of airlines, or go to a service like Expedia,
which aggregates the information, allowing easy comparison of products
and prices. In this way, search costs are lower than if the person has to go to
the high street to find the airlines or travel agents to collect prices.
A person can also reduce bargaining costs, getting more information on
the prices asked by competitors, the reputation of the seller and the quality
of the products and services they are interested in buying. This reduces
the risk of opportunistic behaviour in the negotiation. For instance, a
buyer can go online and check the rating of a seller, customer feedback
on e-commerce services such as eBay or Amazon. A buyer can also go on
Facebook and online forums to check what the previous customers are
saying about the company and its products and services. The buyer can
also check online the sort of contract the company offers, to see that the
terms are acceptable, compared with the terms of other sellers. On the
other hand, the seller can go online to check whether the buyer is genuine
and has a good reputation. Naturally this second option is more difficult
as the internet does not have much information about individual buyers.
However, if the buyer is a company, it becomes easier also to check the
reputation of the buyer. Some platforms are adding the rating of users.
For instance, Uber allows the driver to rate the passenger, as they allow
the passenger to rate the driver. In platform businesses, it is important
to have the feature of rating both agents (sellers and buyers) to reduce
the transaction costs. In the example of Uber, drivers do not want to have
problematic passengers.
103
Lastly, the internet has enabled tools for reducing policing and
enforcement costs. This has been possible especially with the growth of
social media, on top of the internet. If a contract is respected and both
parties are happy, the policing and enforcement costs are minimum,
mainly reduced in the work of checking whether the expected product or
service has been delivered within the expected quality. In the past, when
this did not happen, there will be a private dispute between the parties.
A customer would never know what the other customers were saying
about the company. Nowadays because of social media, people can publish
online their disappointment in relation to companies and products. As this
affects the reputation of sellers, companies have to do their best to try to
sort out the problems with customers as soon as possible. In sum, policing
and enforcement costs are reduced too, considering buyers and sellers
have access to the internet and interactive interfaces such a social media.
Transaction costs can suggest one of the main reasons digital
infrastructures have become so pervasive. And the more apps are built on
the top of each layer of a digital infrastructure, the more we can foresee
these transaction costs going down. For instance, the British payday
loan company Wonga (out of business since 2018) created an algorithm
to check online the probability a person could payback a small loan
for a short period. Using online public information about individuals,
the company reduced the risk of default by checking the capacity of
individuals to pay for their loans. Although the company had controversial
practices of high interest rates, this is an interesting example of a company
trying to use public information about individuals to make decisions on
whether the person is a good or bad risk for a contract, thus reducing their
transaction costs.
Activity 9.2: Applying TC on social media

Drawing on the definitions of the three types of transaction costs and the examples
cited above, go to Facebook or other social media you are familiar with and find
examples of how the service supports the reduction of search, bargaining and policing
and enforcement costs. Be specific in linking the reduction of specific costs to particular
features.
9.2.2 The digital goods characteristics

Economists have struggled to understand the nature of digital goods
and the economic implications of more digital goods and services in
the economy. This is not a trivial question. Economic theory is based on
assumptions about scarcity of resources, from commodities, labour and
capital to final goods and services. It is because there is scarcity, a cost
of raw materials and of production, that there is limited supply to any
physical good. This means that it is possible to put a price in the good
using the market mechanisms such as highest bidder. At a lowest extreme,
the price of a product will tend to be the marginal cost of producing it: in
principle, companies would not produce goods with selling prices below
their marginal costs (although there are economic circumstances in which
this may happen). There are other material goods which are regulated in
a different manner, for instance, when citizens pay taxes to a government
which is responsible for providing everybody with roads, electricity and
public hospitals – so called public goods.
Digital goods, however, might be seen to invert the logic of material goods
for the simple reason that copying a digital good costs almost nothing
in terms of ‘production’ or raw materials. What is the cost of copying an
104
electronic book a million times for a million readers around the world?
The answer is that the cost is so low that we can almost ignore it. In
other words, the marginal cost of production would tend to zero (or be
very close to zero) and thus the price will tend to zero too, following the
mainstream economic perspective of relating prices to marginal costs in
a competitive market. Think of music for example. In this situation, the
problem is that there are no incentives for the producers of such a good.
This was not the case if a million copies of a paperback book were printed
and distributed across the world, or an old-fashioned music CD.
Recently economists have been developing new theories to understand the
difference between physical and digital goods and how goods are wrapped
up in services – think of a plastic DVD as a product (physical good) vs the
Netflix streaming service. They are also interested in how we can ensure
incentives for companies to keep innovating and selling digital goods
despite risks, such as piracy and copying, which may drive the price down.
The next paragraphs summarise the properties of digital goods, following
mainly the work of Danny Quah (2002). The idea is not to explore all
possibilities for defining digital goods, considering the limitations of this
subject guide, but to give you a flavour of the discussion, which is going
to be useful later to understand the discussion on digital products and
services in Chapter 14.
Nonrivalry: Digital goods are nonrival, meaning that when a person uses
it, there is no degradation of the good for another person to use too; there
is no difference between the original and the copy. By contrast, think about
food: if you eat an apple, the second person cannot eat the same apple.
‘Ideas’ are in this sense nonrival, but we do have ways such as patents
and copyright to change this – see below. A paper book in a library can
be borrowed by one person or another, but two people cannot borrow the
same book at the same time. A copy of a digital book can be accessed by
many people at the same time. Each person can have a copy of the original
one, without disturbing the reading of others. And this for a cost which
is close to zero. You can easily think of other digital goods or services like
this such as ideas, computer software, photos and videos, films, music,
newspapers, electronic games, online educational material or artificial
intelligence-based medical advice services.
Infinitely expansible: Digital goods may be quickly copied providing
many with a copy at almost no cost. Although producing the first
copy may be expensive, the reproduction is not. Again, ideas have this
characteristic of being infinitely expansible, being able to be thought and
used by an infinite number of people at any given time. There is a specific
characteristic with this feature because the price of a product which can be
reproduced infinitely tends to go to zero. However, there are some costs
for producing such products. In a situation in which the reproduction of
the digital good is not regulated, innovators would not have incentives
other than intrinsic motivation to create new products, as they would not
be able to sell the product for a good price. Two common solutions here
are found in the laws for patents and of copyright which gives a legal right
to restrict and put a price on copies being made – even if this not always
successful in the end. The open-source software industry, in contrast,
relies more on intrinsic motivation to make people want to contribute to a
product that will be freely distributed.
Aspatial: Digital goods are nowhere and everywhere at the same time
– they are highly and swiftly transportable. For sure digital goods are
inscribed into servers and physical devices and in this sense one can see
105
their presence in a particular space. But here Quah is concerned with the
nature of the digital good being communicated. More directly we can see
that while shipping physical goods around the world takes time and costs
money, digital goods move easily, rapidly and cheaply.
Recombinant: Digital goods may be recombined in different forms, in
such a way that original works can be generated at a low cost. YouTube
has plenty of such examples, in which people create videos from films and
songs, for instance. Another example is the use of digital maps (such as
Google Maps) to pinpoint a particular address and distribute the image
online. Recombination of contents enables high levels of creativity, as the
recombination of content is limitless (see also the concept of generativity).
Activity 9.3: Understanding digital goods

The definitions above characterise the distinctive character of digital goods in relation to
physical goods. Drawing upon the properties of digital goods, analyse the implications of
these characteristic for:
• crypto currencies (e.g. Bitcoin)
• digital maps
• digital legal advice services.
For each of these digital goods, think about nonrivalry, infinite expansible, aspatial and
recombinant.
Economists are concerned about the characteristics of digital goods in

part because it is necessary to find ways of rewarding the producers
and innovators with incentives to create new goods. What would be the
incentive for a musician to create songs, or a music company to sign up a
new singer, if anyone can download music free from the internet? We have
seen this discussion before when recording companies took legal action
against peer-to-peer services such as Napster.
As we know, producers and innovators can protect their interests using
legislation on intellectual property rights (IPR), such as patents, copyrights
and trademark protection, among others. This type of legislation creates
incentives for producers, mainly by stopping people from copying and
spreading digital goods. This solution is a way of fostering innovation
through guaranteeing the payment of economic rents to the producers
of digital goods. However, it also brings a level of inefficiency for society,
as the wide use of digital goods for a very low price does benefit society,
once the cost of the innovation has already been incurred. The protection
of digital goods through IPR also reduces the potential use of them for
recombination, thus limiting the potential number of innovations, relative
to what could be created otherwise.
There is a social tension between the interests of producers and consumers
of digital goods. The question is thus to find the ideal level of protection of
intellectual property rights for each product, considering on the one hand
the incentives for fostering innovation (economic rents) and on the other
hand the benefits for the whole society from having easy access to the
reproduction of digital goods.
Quah suggests two solutions for this conundrum: the use of procurement
and patronage. In the procurement stream, a government or
benefactor pays a reward to a pool of innovators to find a solution to
a problem. The solution is then distributed to the whole society. This
is a solution which is institutionalised already: the military purchase
innovation in this way. In the patronage stream, a government or
106
benefactor would give research awards to competent professionals,

allowing them to decide which innovations would be more relevant. This
model already exists, for instance, when government research councils
give money for pure research in universities, allowing the researchers’
peers to decide on the best innovations on offer. These two models
provide incentives for the innovators at the same time as protecting the
best interests of society at large. They also limit the economic rents of
innovators, creating new incentives for them to look for more innovation.
An additional perspective to protect the value of digital goods is of
associating an NFT to the digital good. The NFT stands for non-fungible
token. The NFT is a unique code, which is associated with a digital
good (e.g. pictures, films, images, audios etc.). It certifies that a copy is
authentic (instead of being a pirate copy), because each NFT certifies one
copy only. It can also work for proving ownership of a digital asset, again
because it provides a unique code associating the digital asset (perhaps
a drawing or a music track) to a particular owner. The control of unique
identifiers and of ownership can be done through a blockchain technology,
which has a high degree of reliability. Blockchain technologies rely on a
peer-to-peer network, so for changing something in a fraudulent way, the
agent needs to have control of the whole network, which is unlikely (not
impossible!) for large networks.
In the examples above, we suggested that a digital good could be copied
at a marginal cost close to zero. Then anyone could make copies and
benefit from the copies, independently of the wish of the producers or
creators of the digital good. The NFT is added as another layer to protect
the intellectual property rights of the producers or creators of digital
goods. Because the NFT is unique and cannot be copied, it is possible to
reduce the legitimacy of illegal copies of digital assets and it is possible to
prove the ownership of a digital asset. Suppose, for instance, an artist has
created a digital painting and puts this digital asset for sale as a unique
piece. The value of the asset is now that there is only one copy of this
asset. To guarantee that there is only one copy, this asset is associated to
an NFT. From that moment, it is known that only this copy with the NFT
is legitimate and any other similar image, even if perfectly good, would
not be a legitimate copy because it will not have this unique identifier
provided by the NFT. The NFT technology has thus generated the scarcity
which is prevalent among physical goods. If someone wants to have a copy
of this digital good or to use the digital goods, they need to pay the price
which is proper for assets which are scarce.
The NFT technology can also guarantee ownership. The same code which
identifies the piece as being authentic may identify who is the legitimate
owner of the piece in any given moment. Controlled by blockchain
technology, this means that a person who copies would not have the
means to change the blockchain, to steal a digital asset from its owner.
The same technology allows, in addition, a trail of the ownership of the
asset, from its creator to the final owner. Thus, the NFT technology can
empower the creators of digital assets to control the availability of their
assets, providing some guarantee for buyers as to the scarcity of the asset
and the ownership of the asset. Naturally, for this NFT approach to work,
the underpinning blockchain technology should be reliable and the control
of process should be managed by a reputable organisation.
9.2.3 The institutional perspective

The institutional perspective is relevant to many aspects related to
information and communication technologies as well as infrastructures.
107
From the perspective of this subject guide, institutional concepts are useful
for understanding the role of digital infrastructures in contemporary
societies and the reason why they have become pervasive. The institutional
view adds a different perspective somewhat different to the economic
ones. You can read more in Scott (2001) and Silva and Backhouse (1997).
The institutional perspective focuses on social structures, which can be
interpreted as the way people behave in a particular circumstance. These
social structures may be supported by many social elements. For instance,
rules and norms may enforce a particular sort of behaviour. In addition,
technology may also be an instrument of enforcing a particular behaviour.
On the other hand, the beliefs and cognitive frameworks which are
internalised by people also affect the way designers, developers, sponsors
and society as a whole understand what a particular technology can do
and how it should evolve. The interplay between technology and society is
intermediated by institutions.
From the institutional perspective, we can derive three conclusions which
apply for digital infrastructures. First, institutional forces influence the
way digital infrastructure is designed and organised. Second, digital
infrastructure influences the way people behave. Third, the adoption of
digital infrastructure is influenced by the collective interpretation of a
variety of social actors. These three statements are explained below.
Institutional forces influence the way digital infrastructures
are designed and organised: As we know, infrastructure is not
developed out of the blue, only because of someone’s creative mind.
Innovators may create any sort of technology they wish; however, if an
infrastructure is going to thrive, it is because institutional forces are able
to influence the structure and architecture, in order to get the outcomes
desired by social actors who represent institutional forces. From this
perspective, we can study the development of digital infrastructures taking
into consideration the aligned interests of those who have fostered the
development of particular services, protocols or applications, rather than
others.
Digital infrastructure influences the way people behave: This
means that digital infrastructures are not neutral, in the sense that the
design/evolution process inscribes particular features and that these
features frame the way people use the infrastructure. For instance,
Facebook has a particular way of organising information. If you would
like to organise the information in a different way, the tool is not going
to be helpful. Facebook decides through an algorithm who is going to see
your posts. It is not possible for you to define either who is going to see
your post organically or what you are going to see on your homepage. Any
piece of technology has this characteristic for enabling some actions and
constraining others. Facebook promises to show you only relevant things
of interest to you, but the cost of this benefit is that Facebook also decides
what is not relevant for you (you may disagree with the algorithm’s
choices). Overall, Facebook can influence what you see and thus the way
that you behave.
The adoption of digital infrastructure is influenced by the
collective interpretation of a variety of social actors: This
means that the diffusion of use of a particular infrastructure does not
depend only on the technical characteristics, or of the economic benefits
it can bring. From this perspective, the most important influence is the
way a relevant group of social actors, such as academics, consultants,
influencers, vendors, consumers, journalists etc., are able to form a
108
coherent view to interpret the infrastructure – its potential and purpose.

When this collective view is positive, then society broadly is going to
adopt that infrastructure. In this view, the social interpretation of the
relevance and forms of use of any piece of technology is the key aspect to
understanding whether it is going to succeed. See more on this topic in
Swanson and Ramiller (1997) (in particular their definition of the concept
of organising vision).
Activity 9.4: How institutions affect the use of social media

Drawing upon the discussion on institutions, analyse how your use of social media is
influenced by the way other people, and particularly your peers, are using social media.
Think first about the social media channels you are using, instead of others you do
not. Then think about the language you use, the sort of content you publish and the
sorts of interactions you have with your friends and other people online and how these
behaviours vary when you are using different social networks (e.g. Facebook, WhatsApp,
LinkedIn and Twitter). Compare your understanding with the impressions of your
colleagues. What is your conclusion on how or to what extent institutional forces such as
rules, norms, beliefs and identity are influencing your behaviour on social media?
9.2.4 The resource-based view (RBV)

The resource-based view of the firm (RBV or RBVF) (Barney, 1991;
Teece and Pisano, 1994) is a theory which focuses on how companies
can use their resources for obtaining sustainable competitive advantage.
Departing from a simple managerial or leadership perspective (e.g. that
leaders have vision and plan future business success), the theory proposes
rather that a company must take into consideration the environment in
their industry and markets and focus on developing superior relevant
capabilities and resources. These then can be a solid basis for sustainable
competitive advantage. Competitive advantage helps a business obtain
better profitability and this helps the business to accumulate and refine
its capabilities and resources for the next cycle. The key is developing and
nurturing appropriate resources, not vision or plans alone.
The RBV suggest that advantage comes when resources are valuable and
rare and difficult to imitate or to substitute. Resources can be all sorts
of assets (buildings, IPR, finance, partnerships, human capital) but also
capabilities, processes, information and knowledge that a company has to
differentiate itself. Such valuable, inimitable and rare resources may be in
particular technologies, in their workforce and skills set, their partnerships
or in their branding. A distinctive combination of relevant resources with
these characteristics can help a company to achieve competitive advantage
in relation to competitors even if individual resources do not have all the
characteristics.
This strategic view proposes that a firm must work on the means to reach
sustainable competitive advantage. The development of the right mix of
resources will not come by chance, of just by design, but by attention over
time to refining the resource portfolio. A firm must analyse its market
and competitors, looking for the types of positioning (price, quality/
differentiation, innovation, service, techniques, partners etc.) which can
help the company to reach the objective of competitive advantage. If the
process of resource development is well established then the advantage is
likely to be more sustainable.
From a digital infrastructure perspective, RBV can suggest to managers
and strategists the best way of adopting digital technologies and digital
109
infrastructures. If a company is adopting any digital resources in the

same way as its competitors, the digital resources are unlikely to have
the characteristics of being rare or difficult to imitate or substitute.
Lots of investment in digital resources and infrastructures are lost
because businesses copy others and follow buzzwords from experts,
instead of deeply understanding their own industry, their own strengths
and weaknesses, as well as those of their competitors. We learn from
the RBV that investments in digital infrastructure, as with all other
resources, do not pay off automatically, if the investments are not driven
by a strategic perspective of building the right mix of resources and
capabilities. A development of this perspective is the concept of dynamic
capabilities (Teece and Pisano, 1994), when the firm creates processes
for constantly improving their competences in various areas. A process
for building dynamic capabilities starts with a good capacity of sensing
the environment to discover opportunities for change and for refining the
resource base so as, in turn, to improve and innovate products, processes
or business models.

This chapter introduced four theoretical perspectives which are relevant
for the study and understanding of digital infrastructures. Each of these
theories help to have a broader understanding on the reasons digital
infrastructures have spread so pervasively, presenting arguments drawing
upon economics, institutional theory and the strategic perspective of the
resource-based view of the firm. The chapter also presented an economics
perspective on the nature of digital goods, showing that society may have
need of models other than the IPR to foster innovation and allow more
people can benefit from digital technologies and infrastructures. These
frameworks will be useful to understanding the next chapters, when we
discuss in more detail particular infrastructures and how they may evolve
in the future.

• explain how to use transaction costs to interpret the diffusion of digital
infrastructures
• apply the basics of the transaction costs perspective to understand how
particular digital infrastructures help suppliers and buyers to better
coordinate the exchange of goods and services
• apply the basics of digital goods perspectives to better understand the
phenomenon of digitalisation of goods
• outline basic ideas of institutional theory and apply them to digital
technologies.

1. Summarise the key concepts of the Transaction Cost perspective.
Analyse how the internet can help reduce the transaction costs of
suppliers and customers. Give examples.
110
2. Do you believe that society needs to find new ways of diffusing digital
goods in order to get more benefits for us all? Suggest four examples
of digital goods or services that might be better developed through the
means of procurement or patronage.
3. Drawing upon the basic ideas of institutional theory, explain how this
theoretical perspective help you to understand the diffusion and reach
of social media services such as Facebook.
111
Notes
112
Chapter 10: The world wide web and web engineering
Chapter 10: The world wide web and web

engineering
Chapter outline
This chapter turns attention to the world wide web. It introduces the
web in terms of its history and development and its place as the world’s
primary information infrastructure. It is sometimes referred to as the
internet’s ’killer app’, the one application that made vast numbers of
people want to be users of the internet. The chapter considers the
development of the web from Web 1.0 – a static web – to Web 2.0 – a
dynamic and transactional web. In the context of Web 2.0, the chapter
considers the new technical discipline of ‘web engineering’ and contrasts it
to older and more established ideas of ‘software engineering’.
10.1 Introduction
The world wide web (WWW or just ‘the web’) is an information
infrastructure that runs on the internet (e.g. in layers on top of TCP/IP). It
in turn underpins many of the most important, widely used and influential
digital services in the world, for example most e-commerce uses the web,
all manner of reference sources, government services, online education,
video sharing, social media and many more.
The web started out from a simpler ambition, to make the information in
documents stored on computers available to people across the internet.
From the outset the web was intended to allow access to many kinds of
digital objects in many formats (text files, PDF files, spreadsheets, audio
and video files, pictures etc.) and at many locations. The primary type of
object that can be accessed using the web is the web page.
A web page is a kind of digital document that normally contains both
information (e.g., text) and links to other web objects (hyperlinks).
Indeed, the core innovation the web brought when it was first developed
in the late 1980 and early 1990s was to allow any given web page to
contain hyperlinks that lead on to other documents and pages. A hyperlink
is a direct link in one document or web page that embodies an IP address
in the form of a URL. Click on or select that link and you move on to the
referenced web page. We are very familiar with these hyperlinks in web
pages, seeing them as either explicit and clickable URLs e.g. https://
en.wikipedia.org/wiki/World_Wide_Web, or as embedded in a clickable
word, phrase, picture, or icon.
By using hyperlinks there is no prescribed or hierarchical set of files,
folders or directories that need to be traversed to go from one page/object
to another – the web is not like a neatly and consistently organised filing
cabinet where specific information is located in a pre-designate place.
Rather, the links are created by people as they come to understand their
particular needs. It is the user/author of any given page who sets and
revises the hyperlinks and more who generally can update a page as and
when they want. Information is more dynamic and able to be reconfigured
as and when it seems appropriate.
The implications of this freedom for authors/publishers and for how we
access and present information has been profound. Now everybody can
be an author and (potentially) reach almost everybody else in the world.
113
The digital economics of the written word, music and visual materials
has profoundly changed too. The amount of information in circulation in
the world has exploded and much of this is now accessed ‘for free’ – for
example do you buy a newspaper – or do you access news ‘for free’ –
leaving aside the adverts?

• Introduce the web as the information infrastructure underpinning
one of the earliest information services that drove the uptake of the
internet for mass users.
• Present the web’s basic architecture as a client-server system that
uses its own HTTP communications protocol and the HTML mark-up
language.
• Chart the evolution of the web from a static Web 1.0 to a dynamic and
transactional Web 2.0.

• describe how the web was created and evolved
• identify the key characteristics of the web basic architecture
• differentiate between Web 1.0 and Web 2.0 functionalities and
applications
• apply in basic-level HTML for producing content for the web.

[ISBN 9781292417752]. Chapter 7, Sections 7.2 and 7.3.
Berners-Lee, T. ‘The world wide web: past, present and future’, Draft of a paper
for IEEE Computer 1996. Available at the World Wide Web Consortium
(W3C) website: www.w3.org/People/Berners-Lee/1996/ppf.html
Berners-Lee, T. ‘The future of the web’, (2007). Testimony before the United
States House of Representatives Committee on Energy and Commerce
Subcommittee on Telecommunications and the Internet. http://dig.csail.
mit.edu/2007/03/01-ushouse-future-of-the-web.pdf
O’Reilly, T.P. ‘What is Web 2.0?’ (2005) www.oreilly.com/pub/a/web2/archive/
what-is-web-20.html

‘A little history of the world wide web’. Available at the world wide web
Consortium (W3C) website: https://www.w3.org/History.html
‘The web’ in Ryan, J.(ed.) A history of the internet and the digital future.
(London: Reaktion Books, 2013) [ISBN 9781780231129].
Wikipedia page ‘Web 2.0’: https://en.wikipedia.org/wiki/Web_2.0

This chapter introduces the history of the web and how it has become
the primary information infrastructure of our times. It explains the work
of Tim Berners-Lee and how it has evolved to create the current forms of
presenting content online. The chapter discusses the differences between
Web 1.0 and Web 2.0 and introduces the basics of HTTP/S (HyperText
Transfer Protocol) and HTML (HyperText Markup Language), as well as
showing how pages are found on the internet. Finally, it introduces in basic
level the logic of web search and the concept of web engineering.
114
10.2 Chapter content

The world wide web is a digital infrastructure that supports a set of global
information services. Well-known examples of such services are Wikipedia,
YouTube, online newspapers and social media sites like Facebook. It is an
infrastructure and serves as a shared resource that underpins the provision
of all manner of information services to billions of individual users who
provide (post) and request (browse for) information every day. In his early
writing Tim Berners Lee, the web’s founding father, describes it as a ‘global
information system’ (Berners-Lee, 1996). But infrastructure is probably a
better or more accurate word to describe it because of its emergence and
development over time and its role as a standardised service to support
multiple different users and uses. In any case, from now on we will just
call it ‘the web’.
10.2.1 World wide web architecture and protocols

Like any other infrastructure the web has its own architecture – described
as a client-server architecture and its own protocols. And, of course, this
architecture is layered – layered on top of the internet and layered with
the web as extra functionality is added ‘on top’ of basic functionality – for
example as security is added on top of basic message transmission.
The web’s architecture is based as on two types of Hosts – clients and
servers, linked together by the internet so of course using TCP and IP.
Clients are computers that request information and servers supply or
serve it. My computer is a client computer when I browse the web and
depending on what I click on in a document or a web page my computer
then exchanges messages with the associated server to gather (Get)
information from it. Both servers and clients run software that supports
their role. Client computers run client software, such as Edge, Safari,
Chrome or Firefox. Server computers, which hold web pages and other
information resources, use server software. Indeed, the majority of web
servers across the world use just one particular type of server software –
the Apache HTTP server. This is an open-source product, freely available
and maintained by volunteers in a collaborative effort.
The web uses the internet as its means of passing messages (data).
Then, on top of TCP/IP, the web has its own specific protocols and data
structures to support its requirements. Two of these protocols that you are
probably a bit familiar with are the protocols HTTP and HTTPS. HTTP
stands for HyperText Transfer Protocol and is the protocol that takes a URL
(e.g. as typed in your browser or clicked in a page you are viewing) and
sends a request message across the internet to the indicated Host server (a
GET message). In response the server sends back the indicated web page
which we hope contains the information you want.
So, if I click on www.ibm.com a Get message is sent to IBM’s web server
(a Host on the internet) and the text of their homepage is sent back to my
client software (my browser). I then see it on my screen. IBMs homepage
is full of text, images and other hyperlinks/URLs so I can navigate onwards
into the IBM website to find what I want by clicking of these hyperlinks. I
could also use the ‘Search Box’ on the screen to find specific items in IBMs
website (we discuss more general search engines like Google below).
In the familiar jargon we ‘surf’ the web, an ocean of information. When I
click on www.ibm.com, I ‘go to’ their website, but it would be a bit more
correct to say that it (or a bit of it) ‘comes to me’, thus the text and the
layout information including pictures (e.g. a web page) is sent back over
115
the internet to my client computer and my client software. HTTP and

HTTPS are the protocols that manage this request and send traffic.
So, what is the difference between HTTP and HTTPS as protocols? HTTPS
does in general the same job as HTTP but the added S indicates Secure
– e.g. that the traffic the protocol passes to and from is encrypted. Some
older sources describe the S as meaning ‘Secure Socket Layer’, a separate
protocol for encryption, but this specific protocol is often not used today,
being superseded by other better ones.
Server Side Client Side
Web content
Web page, pdf file,
photo, video etc
Protocols: Application
HTTP, HTTPS etc.
Web server software Content: Web browser
Apache, IBM HTTP server, HTML, Javascript etc. Chrome, Firefox, Explorer
Tomcat etc. Safari etc.
Host computer Internet Personal device

IP address (TCP, IP) PC, laptop, smartphone etc.
Figure 10.1: Client and Server interface.

There is one last bit of this (rather simplified) model that we need to add.
When the IBM homepage is sent to me it is sent as some raw text and
images plus some extra information to say how these should be displayed.
So, to grossly simplify, when I click on www.IBM.com and after HTTP
has done its job of a GET and I see the homepage with the words ‘News:
IBM Reveals New AI and Cloud Powered Fan Experiences for Wimbledon
2022’ in the centre of the screen (accessed July 2022). These words are in
big silver letters, but when sent to me the text is sent separately from the
instructions as to how large the letters should be and in what colour. The
words and then how they should be displayed their size, the colour, font,
positioning and perhaps associated hyperlink, are all coded separately in
the HTTP message that arrives at my computer.
It is then my browser that puts it all together (renders it) on the screen.
If you have a different browser on a different device (say a smartphone),
then the way it looks may well be different. Remember too, if you ‘go
to’ www.ibm.com in two years’ time the website will still be there, but
its content will be different. Indeed, this homepage for IBM is probably
updated more than once per day, which is central to the whole idea of the
web as a dynamic and constantly changing information infrastructure.
If I select the text on the IBM homepage, right click and select ‘Inspect’
from the menu that pops up (I am using a Windows computer and Chrome
browser) I see in a parallel window the actual text that IBM’s server sent
back to my browser. It is messy! And I am sure that most of you will
not make much sense of what you see. However, all the codes you will
see are there to convey how the page should look and behave. They are
written (coded) in HyperText Markup Language (HTML). Today any word
processor you use can probably save a file in HTML format (e.g. with
HTML codes) and that file can be opened and ‘read’ by any browser.
HTML is central to how the web works, allowing an author freedom to
show how a web page should look and then allowing the client software
116
(browser) to make the best effort to achieve this. As you may have noted,
this has echoes of how the internet also works on a ‘best effort’ basis and it
allows the two ends of a communication to ‘negotiate’ when things break
down of become difficult – think of TCP.
The use of a client-server architecture for the web and the use of HTML
are both clever choices that distribute the work of getting and displaying
information between the server who will find the page requested and
send it as a standard set of data including text and HTML mark-up and
the client who does the work of rendering the data for the screen of the
specific device as best it can. HTML is a global standard, so if the server
sticks to HTML then any client machine or software that knows HTML will
work. Again, we might see this as a return to the logic of the early era of
the internet when architecture was found that could support all kinds of
machines and adapt to them as they developed and grew. We should also
note that HTML has gone through different versions over time, but these
are generally ‘backwards compatible’ so using a new version does not
make all websites using older versions inoperable.
Figure 10.1 is a simple depiction of a basic web architecture (its structure).
It is, however, limited to the model of browser (with human user) and
server (machine with web pages). But we could expand this figure in two
ways. First to include the search engines that are a central part of the web
and second to include the many machine-to-machine interactions that the
web supports (including the search engine web crawlers). If you were to
redraw the figure including these elements, how would you do this? More
generally when we think about the architecture of any infrastructure we
need to look at the fundamental relations of the element involved and
that usually means looking into the variety of exchanges, processes and
functionalities.
Activity 10.1: Working with HTML

Prepare a fairly simple Word document that uses two fonts each in two different sizes and
colours. Make it short, about four lines of text.
Save it as an .HTM document e.g. as a ‘web page’.
If you do not use Microsoft Word then you preferred word processor will almost certainly
be able to save a document as an HTML file.
Then (assuming you use Word), use the Word ‘View’ menu to look at its HTML code by
selecting ‘HTML Source’, near the bottom of the menu.
You will find your text and the codes that show font, size colour etc., towards the end
of the file after a lot of ‘preliminary’ code. How much can you work out of the basic
grammar of HTML from this exercise?
Please note that you are not required to have any skill in reading or writing HTML for
this course, but you should understand what a mark-up language does and why it is an
important part of the web’s architecture.
10.2.2 History of the world wide web

The origins of the web are associated with one person in particular, Tim
Berners-Lee. In the 1980s he was a computer scientist working at CERN
– the large European particle physics research laboratory on the border
of Switzerland and France. He developed an early ancestor of the web
called Enquire to meet some local needs for documenting the extensive
IT facilities of CERN. After some modest success he developed his ideas
further into the form of the web we see today. That is of documents
(pages) that contain hyperlinks that lead to other documents (pages). It is
117
as simple as that. Of course, choosing to make the URL hyperlinks a kind

of IP address meant that any page could link to any other page anywhere
on the internet – anywhere with an IP address and an internet connection.
Still, it was not an immediate success. In Chapter 8 of Ryan (2013) you
can read about the ‘distinct lack of interest’ that there was for the web
at CERN. Indeed, Berners-Lee continued on with the project despite his
bosses shelving the project on two occasions. The first ‘browser’ was
written by an intern on a short posting at CERN. The major breakthrough
(!) came when CERN was persuaded to put their telephone directory
on the new web as an HTML document – not the most exciting project
perhaps, but a foot in the door.
The web really started to take off when more browsers were written,
browsers that could run on common computers of that time. It was
not until 1993 that browsers (client software) were available for most
common desk-top computers – the last one being for the IBM PC – called
Cello. But the big commercial success was a browser called Netscape
which emerged from a US government research centre, The National
Centre for Supercomputing Applications (NCSA) and their early browser
Mosaic. The Netscape company was one of the first ‘internet’ start-ups that
made the big time, raising about $1.4bn when it went public on the stock
exchange in August 1995, with its value rocketing to almost $3bn in the
next few months. Netscape was perhaps the first pure ‘internet’ company.
10.2.3 Web 1.0 to Web 2.0

Web history, since the days of Netscape, is most often told in general
terms as the shift of all kinds of business, administration, science and
entertainment to this new infrastructure. Companies soon had websites
and documents and files of all kinds were distributed by being placed on
the web. Some industries moved fast to the web, music, entertainment,
books (Amazon) and transport (Airlines). Others were slower,
government, retail, financial services, often for good reasons e.g. security
and accessibility concerns.
However, there is a more analytical perspective we should understand,
which is the shift from the early web of static documents in organisations
websites linked by hyperlinks, but with limited ‘search’ capacity, to today’s
web.
The initial web was rather simple and static but over time of course this
changed. The web came to contain much more than documents, sites
were not just for established organisations but for individual people, new
start-ups and many community and social groupings. There also grew up
a vibrant industry performing web searches. At first these would be sites
(such as the original Yahoo) that kept list of ‘good’ or useful websites – a
kind of directory. But over time this approach has given way to search
engines, like Google or Bing, which perform dynamic searches and
indexing on the web and on this basis allow people do their own ad-hoc
searches. The way that Google and other search engines work is complex
and involves continuous web ‘crawling’ to obtain information for a huge
index of web content. One other key aspect of a search engine is not only
to index web pages according to their content, but also to ‘rank’ them by
how many other pages link to them. A page that contains content on topic
x and is referenced by many other pages is judged to be more interesting
and is thus shown in at the top of the list a web search.
118
Activity 10.2: Google’s page rank algorithm

Research and write a short report on how Google’s ‘page rank’ algorithm works and thus
how Google decides which websites and pages to show to you when you make a search.
In the report explain how websites and pages are indexed (web crawlers) and how they
are ranked.
Also research how paid advertising is incorporated into Google pages. Remembering that
most of Google’s profits are made by ‘serving’ advertisements to people like you and me
– partly on their search pages and partly on other pages we surf such as newspapers or
business websites
Most search engines are fairly literal, even a bit dim. They take words
and phrases and index the web on this basis. But there are many people,
including Tim Berners-Lee the web’s originator, who are trying to create
a different kind of web – the semantic web – which has a better sense
of what things on the web ‘mean’ than is possible from text strings. For
example, in the case of a video on the web, or a picture, or a poem it is
quite difficult for a program to analyse ‘what it is about’, other than from
the title. But in a semantic web, with the associated infrastructures, this
might be possible and take us closer to the web as a global ‘brain’. (Do you
think this is possible or desirable?)
The shifting and evolution of the web is often summarised in the idea
of an early, simpler, static Web 1.0 and a new, different, ‘upgraded’ Web
2.0. Of course, there is no one moment when it shifted and it is not really
about the web changing in technical terms (though that does happen –
infrastructures evolve), but rather our ideas of what we could do with it
changed. Ideas of Web 1.0 and 2.0 have been around from about 2004
(read O’Reilly, 2005 and the Wikipedia page on ‘Web 2.0’). And some
people want to move on the debate to Web 3.0 or 4 and so on.
There are some changes that we can identify in how the web is
experienced and used by people which add up to Web 2.0.
For example, in Web 2,0 people interact more with each other, people
generate more of the content (user-generated content) e.g. in Blogs,
Vlogs or Facebook. Another powerful example of user-generated content is
the online encyclopaedia we recommend in this course, Wikipedia, where
the content is the product of many enthusiastic volunteers. More generally,
using Web 2.0 people can join in specific interest groups (virtual
communities) which may transcend traditional geographic boundaries
and can have interests ranging from politics to stamp collection to 1930s
baseball statistics. People also expect to comment on others’ efforts,
contribute to them or use them for their own purposes, for example in
photo sharing sites like Flickr.
Some of the changes for Web 2.0 are in the services that the web can offer.
In Web 2.0 there are many more transactions on the web. For example, the
kinds of transactions (and security) needed to have electronic commerce.
This includes, of course, financial services that let people pay online and
to help them trust as website they may have never heard of before. More
generally we can say that the Web 2.0 is a web where different services
(or servers) can interact together to deliver new kinds of service for web
users. E-commerce again gives an example of how a website that sells,
say, clothes links to PayPal for secure payment services and to a courier
company to deliver the parcel. It may also ask its customers to comment
on the goods it offers and give positive or negative reviews of specific
products. All this happens on the web, in careful interactions among their
119
separate companies and their clients, each providing separate parts of the
service you want.
10.2.4 Web engineering

The web has become a rather complex infrastructure – at least in technical
terms – and has gone far beyond simple HTTP- and HTML-based pages
of Web 1.0. Today’s websites, full of video, animation, secure financial
transactions, encrypted data and linked to sophisticated and intelligent
services, require a high level of knowledge and skill to produce and to
keep operational seven days a week and 24 hours a day for thousand or
millions of users. Thus, a new discipline of ‘web engineering’ has emerged
to embrace such skills and techniques.
Many organisations today are so reliant on their website that it is as
important and central to their survival than the older ‘internal’ information
systems. If the payroll program fails for a couple of days, it can be
tolerated, but if the website is down for three hours it might cause chaos.
How long could an airline, an ecommerce company like Amazon or Baidu,
or a logistic company like DHL stay in business without a website? And
how long could a country’s government survive if key websites were
inoperative? So, web engineering is not just about knowing how to build
web services well, but also how to keep them running and respond to
problems.
Problems of course come in many shapes and forms, but one class
of problem are those associated with malicious attacks on web
infrastructures. In Chapter 5, we discuss an attack on the internet in
2016 that disabled a part of the DNS. Reacting to such attacks (e.g. being
prepared to move the DNS service to other services), or to maintain
parallel websites on different servers, is part of the remit of web
engineering and a response to the web being increasingly seen as a ‘critical
infrastructure’.

This chapter introduced the history of the web – the internet’s ‘killer app’.
It explained the relevance of the web and how it has become the primary
information infrastructure of our times. The chapter explained the creation
and development of the web. The chapter discussed the differences
between Web 1.0 and Web 2.0, emphasising the relevance of the second
for getting more engagement and participation from users, through the
user-generated content. The chapter also introduced the basics of HTML,
the logic of web search and the concept of web engineering.

• describe how the web was created and evolved
• identify the key characteristics of the web basic architecture
• differentiate between Web 1.0 and Web 2.0 functionalities and
applications
• apply in basic-level HTML for producing content for the web.
120

1. In his testimony to the US congress in 2007, Tim Berners-Lee
(2007) says that the success of the web is built on ‘three critical
factors: 1) unlimited links from any part of the web to any other;
2) open technical standards as the basis for continued growth
of innovation application; and 3) separation of network layers,
enabling independent innovation for network transport, routing and
information applications’. Explain each of these factors and how
they might lead to success. In each case give one or two examples to
illustrate your points.
2. Is it justifiable to describe the web as an information infrastructure?
In your answer assess how well it meets the general criteria for being
considered as an infrastructure as set out in Chapter 2 of this guide.
3. Describe the general distinction that is often made between Web 1.0
and Web 2.0. What three specific features or abilities would you
identify as being most important in identifying what changes in the
shift from Web 1.0 to Web 2.0?
121
Notes
122
Chapter 11: Cloud computing as an innovation in infrastructure
Chapter 11: Cloud computing as an

innovation in infrastructure
Chapter outline
This chapter aims to further develop the concept of cloud computing
(introduced in Chapter 2). We conceptualise cloud computing, explaining
the origins of cloud computing, the reasons why cloud computing has
become a pervasive model for providing computing services, the types
of cloud computing and the risks this digital infrastructure presents. We
emphasise in the chapter the layered architecture of cloud computing,
supported by communication standards, the network effects and
economies of scale of cloud services and some aspects of how cloud
services have enabled innovation in digital infrastructures.
11.1 Introduction
Chapter 2 introduced the idea of cloud computing. We need now to have a
more formal definition of cloud computing, to account for the complexity
of this system. The National Institute of Standards and Technology (NIST,
a US institution), defines the term as follows:
Cloud computing is a model for enabling convenient, on-demand

network access to a shared pool of configurable computing
resources (e.g., networks, servers, storage, applications and
services) that can be rapidly provisioned and released with minimal
management effort or service provider interaction.
(NIST)
This is a short definition which embeds a lot of complexity. NIST defines

that a cloud model should promote availability of resources, through
having an aggregated amount of hardware, software and services which
can be shared and combined by users in accordance with their needs
(resource pooling). This is in line with the layered architecture we
have been discussing for digital infrastructures. All these layers follow
communication standards which allow the cloud services to work as an
integrated system, in spite of being constituted of many parts. As the
offer of cloud service matures, it is possible also to combine services from
different providers, which increases the potential level of innovation
created by the cloud infrastructure.
The users define the number of services they want to buy at any time (on-
demand self-service). For this to happen, the resources should be available
through a broad network access and the cloud service should be able to
provide the number of services demanded, thus having a rapid elasticity
of the provision of services. Finally, as the services are on demand, the
provider should be able to measure the amount of service used, in order to
bill the client properly only for the resources they have used.
As with the other digital infrastructure discussed in this guide, cloud
computing has evolved through a long process and it still has a long way
to go in terms of becoming more pervasive and enabling more businesses.
123
Activity 11.1: The technical characteristics of cloud computing

Drawing upon NIST’s definition, identify and explain the key technical characteristics of
cloud computing services. You may benefit from consulting further online NSIT documents.

• Conceptualise cloud computing.
• Explain the origins of cloud computing.
• Discuss the advantages and challenges of cloud computing.
• Link the discussion of cloud computing to key concepts of digital
infrastructures.

• describe the concept of cloud computing and the underlying
assumptions behind the services
• compare the advantages and challenges of cloud computing to inform
decision making
• explain the connection between cloud computing and digital
infrastructures
• applying basic cloud computing to business innovation.

Armbrust, M., A. Fox, R. Griffith, A.D. Joseph, R. Katz, A. Konwinski, G. Lee, D.
Patterson, A. Rabkin, I. Stoca and M. Zaharia ‘A view of cloud computing’,
Communications of the ACM 53(4) 2010, pp.50–58.
[ISBN 9781292417752]. Read all the parts of the book related to cloud
computing. See in particular Chapter 5.

Badger, M.L., T. Grance, R. Patt-Corner and J.M. Voas ‘Cloud computing
synopsis and recommendations’, NIST. (2012) www.nist.gov/node/588576
Campbell-Kelly, M. ‘Historical reflections: the rise, fall and resurrection of
Software as a Service’, Communications of ACM 52(5) 2009, pp.28–30.
Carr, N. The big switch – rewiring the world, from Edison to Google. (London,
New York: W.W. Norton & Company, 2008) [ISBN 9780393062281].
Cusumano, M. ‘Technology strategy and management: cloud computing and
SaaS as new computing platforms’, Communications of ACM 53(4) 2010,
pp.27–29
Fingar, P. Dot.cloud. The 21st century business platform. (Tampa, Florida:
Meghan-Kiffer Press, 2009) [ISBN 9780929652498].
Mulholland, A., J. Pyke and P. Fingar Enterprise cloud computing: a strategy
guide for business and technology leaders. (Tampa, Florida: Meghan-Kiffer
Press, 2010) [ISBN 9780929652290].

NSIT: The National Institute of Standards and Technology. https://www.nist.
gov/itl/cloud-computing
Campbell-Kelly, M. ‘Historical reflections: the rise, fall and resurrection of
Software as a Service’, Communications of ACM 52(5) 2009, pp.28–30.
Cloud computing: https://en.wikipedia.org/wiki/Cloud_computing
124
Cloud computing: https://www.nist.gov/news-events/news/2011/10/final-

version-nist-cloud-computing-definition-published
Cloud computing: http://nvlpubs.nist.gov/nistpubs/Legacy/SP/
nistspecialpublication800-145.pdf

This chapter further develops the concept of cloud computing, including
some historic events on how this digital infrastructure has evolved. The
chapter shows the link between cloud computing and innovation in digital
infrastructures. The chapter introduces a typology for cloud computing
services and deployments, discussing the advantages, challenges and risks
of this digital infrastructure. The chapter focuses on the concepts of cloud
computing. Further discussions on cloud computing services are in Chapter
14, when we discuss the delivery of digital products and services.

11.2.1 Origins of cloud computing
Cloud computing or utility computing are relatively new concepts, but
this does not mean that similar conceptualisation of services have not
been there for a long time. For a very short introduction on this evolution
of cloud services, you can read more details in Campbell-Kelly (2009),
focusing on the services offered in the USA. In the paragraphs below, we
summarise key ideas from this reference.
The author says the first cloud service has been offered in 1932, when IBM
opened the first service bureau, offering punched-card Electric Accounting
Machines (EAMs). This service was offered to companies, which would
prepare the punched-cards and bring them to IBM, which would them use
their EAMs to make the calculations.
In 1949, Automatic Payrolls Inc. (later named Automatic Data Processing,
ADP) started offering payroll processing, using both bookkeeping and
punched-card machines. Its business model was very similar to the IBM
service bureau, with a differentiation: the company would collect the
material for the calculations and would return the original sheets and
cards to the customer.
As mainframe computers become available, in the mid-1960s, several
American companies started offering remote access to time-sharing
computers. Companies would pay for use of these resources. In the 1970s,
General Electric also entered the business of data processing, following a
similar business model.
The concept of time-sharing computing declined in the 1980s for the
simple reason that personal computers became available and affordable
for companies and communication costs at that time were high. Thus,
it would be cheaper to buy computers and process data locally than
transporting data to the service provider.
The situation changed again in the mid-1990s. In spite of computing
hardware being much cheaper, cloud type service industries took off.
From a technical perspective, communication networks became more
efficient and cheaper and centralising the management of shared and
scalable servers easier. From a managerial perspective, the maintenance
of computing infrastructures became more complex, more expensive
to maintain and demanding of management attention. Particularly
challenging and expensive are addressing the issues of reliability and
security and individual companies may not have the necessary resources to
manage this complexity.
125
Cloud computing has been an important area thus for outsourcing

operational functions. Outsourcing in this context is the contracting of third
parties for doing specific business processes and operations for a company.
Instead of managing servers and software locally, inside the organisation,
third party companies are contracted to provide these services through
the cloud. For instance, a company may decide not to invest in servers,
preferring to contract Amazon or Microsoft for hosting its data and its
software. In the literature, cloud computing service models (discussed
below) are often discussed in the context of outsourcing, although
outsourcing is a broader concept and companies may outsource all sorts of
operations, from recruitment, security and cleaning to manufacturing and
human resources – some as cloud services and some in other ways.
11.2.2 Innovation in digital infrastructures

For cloud computing to be able to deliver its services, it has been necessary
to reach a level of maturity of the technological infrastructure as a whole.
In other words: the whole technological infrastructure has evolved in
the last decades so as to enable the emergence of cloud computing as we
know it today (and as it may evolve in the future). This idea is aligned
with the layered model (hourglass architecture) discussed in Chapter 6, in
which infrastructures evolve and new layers emerge.
Across time, the internet infrastructure has matured to support cloud
services. The network has become more reliable, with more robust
hardware and software, to reduce the downtime of services in general. This
implies that both the internet backbones and the ISPs (internet Service
Providers) have improved the quality and the resilience of services, thus
being able to cope with diverse problems (such as weather changes and
accidents involving the hardware structure; any sort of vulnerability of the
software layer; and the specific needs of keeping the communication flow
of large amount of data). The internet has also become faster (the spread
of broadband services, available also in mobile devices) and cheaper,
because of new technologies and because of economies of scale and
positive externalities (the more users join the internet services, the cheaper
for each one to have the same service available and more valuable it is).
At a localised level, there were also improvements in the efficiency of
using servers through virtualisation techniques (which allow the same
server to be split into separated servers, through the use of software layers,
thus improving the productivity using the same installed capacity). In
addition, it has become easier to duplicate databases, thus increasing the
security of cloud services (backups) and to change data from one storage
to another, thus allowing further interactivity of different data sets to
deliver a seamless experience to the final user of these services. When one
uses an email or a social media channel, one does not know from where
the data is coming from, and how the data is integrated as one service in
the digital interface.
All these elements together have brought economies of scale, with
reduction of costs. Cloud service companies have been able to build more
efficient data centres, with server farms, saving money from the purchase
and maintenance of servers to the choice of facilities (normally remote
places which have cheaper land) and maintenance of buildings (with less
use of energy for the refrigeration of servers). This economy of scale would
not be possible if the internet has not improved to allow fast transmission
of data from one place to any other.
In addition, to increase operational efficiency, cloud services have
enabled businesses to innovate in terms of new products and services,
126
new business processes and new business models. Adopting a cloud

infrastructure means organisations need less resources to innovate. This
happens because businesses can replace fixed costs (such as investment
in digital infrastructure) by variable costs, reducing the barriers to
entering new markets. This means small- and medium-sized companies
are empowered to challenge incumbents. Instead of investing upfront on
fixed assets, companies may build their business infrastructures using the
scalability offered by cloud providers (paying just for the used resources).
Companies may even combine tools from different providers, using for this
communication standards (interoperability) and rely on the expertise of
the cloud provider to gain competitive advantage.
Activity 11.2: Innovation and cloud computing

Explain how innovation in digital infrastructures have enabled the emergence of cloud
computing as we know it today. Explain how cloud computing is an enabler of innovation
for businesses.
11.2.3 Types of cloud computing

This section summarises the three types of cloud computing service
models – Software as a Service (SaaS), Platform as a Service (PaaS) and
Infrastructure as a Service (IaaS) – and four deployment models – private
cloud, community cloud, public cloud and hybrid cloud. You can find
these definitions on Wikipedia, following the NIST’s definitions (https://
en.wikipedia.org/wiki/Cloud_computing). We further develop details on
each of the service models in Chapter 14.
Cloud computing service models
Software as a Service (SaaS): The capability provided to the
consumer is to use the provider’s applications running on a cloud
infrastructure. The applications are accessible from various client devices
through either a thin client interface, such as a web browser (e.g., web-
based email), or a program interface. The consumer does not manage or
control the underlying cloud infrastructure including network, servers,
operating systems, storage, or even individual application capabilities, with
the possible exception of limited user-specific application configuration
settings. Example: any sort of software service used in the cloud, such as
Google Gmail, Office 365 (Microsoft) and social media tools.
Platform as a Service (PaaS): The capability provided to the consumer
is to deploy onto the cloud infrastructure consumer-created or acquired
applications created using programming languages, libraries, services and
tools supported by the provider. The consumer does not manage or control
the underlying cloud infrastructure including network, servers, operating
systems, or storage, but has control over the deployed applications and
possibly configuration settings for the application-hosting environment.
Example: Force.com platform which supports the development of
applications related to Salesforce.com.
Infrastructure as a Service (IaaS): The capability provided to
the consumer is to provision processing, storage, networks and other
fundamental computing resources where the consumer can deploy
and run arbitrary software, which can include operating systems and
applications. The consumer does not manage or control the underlying
cloud infrastructure but has control over operating systems, storage and
deployed applications; and possibly limited control of select networking
components (e.g. host firewalls). Example: virtual machines at Google and
Amazon.
127
Cloud computing deployment models

Private cloud: The cloud infrastructure is provisioned for exclusive use
by a single organisation comprising multiple consumers (e.g., business
units). It may be owned, managed and operated by the organisation,
a third party, or some combination of them and it may exist on or off
premises.
Community cloud: The cloud infrastructure is provisioned for exclusive
use by a specific community of consumers from organisations that have
shared concerns (e.g., mission, security requirements and policy and
compliance considerations). It may be owned, managed and operated by
one or more of the organisations in the community, a third party, or some
combination of them and it may exist on or off premises.
Public cloud: The cloud infrastructure is provisioned for open use by
the general public. It may be owned, managed and operated by a business,
academic, or government organisation, or some combination of them. It
exists on the premises of the cloud provider.
Hybrid cloud: The cloud infrastructure is a composition of two or more
distinct cloud infrastructures (private, community, or public) that remain
unique entities, but are bound together by standardized or proprietary
technology that enables data and application portability (e.g., cloud
bursting for load balancing between clouds).
Activity 11.3: Identifying cloud computing services

Based on the definitions above, find three examples of each type of cloud computing
service (SaaS, PaaS and IaaS). See how these services are related to each other and find
companies which provide services in these three segments.
11.2.4 Advantages of cloud computing

There are many advantages of using cloud computing and they explain
the reasons why this digital infrastructure is increasingly relevant for all
sorts of organisations and for individuals. In the next paragraphs, we
explain some of these advantages, focusing on the business use of cloud
computing.
Improve scalability: In a conventional data centre, users know the
amount of data they may host. It is a fixed number. Using cloud services,
users have a scalable volume of computing resources on demand: the
greater the demand, the great the use of the system and vice-versa
(elasticity). Although no system has infinite capacity, from the individual
user’s perspective it is like the capacity is infinite. Knowing scalability
is available, executives may spend less time estimating the volume of
necessary computing resources, focusing on other aspects of the business,
such as innovating and becoming more competitive.
Reduce need of investments: In a conventional data centre, the
company would need to invest upfront on the servers, to build the
necessary capacity to attend the final customers. These are investments
in fixed assets. For instance, a developer providing a game in a mobile
operating system would need to buy and maintain the servers to attend
the number of customers the company would expect to have in a period of
time. However, if the game would not be very successful, the investment
in servers would be a waste of resources (the company would have
invested in an infrastructure which is not used). On the other hand, if the
game would be a great success, the company could find itself without the
128
necessary computing capacity to attend all potential customer. This would

be a loss of opportunity, as potential customer could not buy the service
because of the limitation in the computing power.
With cloud services, the developer would buy the access to a flexible
amount of use, thus increasing or reducing the use of the service in
accordance with the demand, reducing the risk of the whole investment.
The services would be paid in accordance with the needs only, thus if more
customers are at your door, you have the means to sell your services and
pay for the cloud service proportionally. If you have less customers at your
door, you do not need to pay for the cloud services you have not used.
Reduce costs: The main driver for the use of cloud services is the
reduction in costs. The economies of scale gained by the cloud service
providers (as explained in the introduction) are passed (at least partially)
to the customers of these services. These are straightforwardly reductions
in the costs. On the top of this, we have another element of speed, which
also indirectly affects costs. Independently of savings costs directly, cloud
services may help companies to gain speed in the provision of services or
in operations. And as speed means more quality and more efficiency, in
any way the company is saving costs and improving profits for speeding
up its processes. The reduction in costs is often cited as the main reason
companies would adopt cloud services.
Improve access to organisational content and operations:
Moving business to the cloud means access to the organisation’s operations
from any place, from any compatible device. People may access the
necessary resources in the company’s site, or from their homes and mobile
devices. This gives great flexibility in the way we do businesses, with
more flexibility not only in the way we work but in the way we provide
services. A consultant, for instance, can have the information about the
client when it is necessary, in the consultancy side, in the client’s side, or
on the road. This architecture also allows better collaboration between
companies and professionals. Think, for instance, about the benefits you
have of being able to collaborate with your colleagues in coursework
assignments using a service such as Dropbox and Google Drive (both in
the cloud). Companies may better create collaborative structures among
different contributors using cloud services, which is also favoured by the
convergence of technologies.
Increase flexibility: Investing in corporate infrastructures is very
expensive. It is an area of risk for any company, as any part of the
investment may be either not enough or obsolete (when a particular
technology is not adequate for a particular business operation). Opting
for cloud computing increases the flexibility for the organisation in terms
of the digital infrastructure architecture to build. Instead of making
investments in fixed assets, the company makes a contract with a cloud
services provider, changing the services to buy in accordance with the
needs. In this way, the risk for investing at least in part of the infrastructure
can be outsourced to the cloud service provider. Increasing flexibility
improves the responsiveness of digital infrastructures to the business,
enabling growth and capacity for changing. Business organisations become
more agile when they can draw upon cloud resources.
Better management of complexity and risk: As digital
technologies expand their pervasiveness in business, the complexity for
managing digital infrastructures is increasing and with more complexity,
we have also more risks. Adopting cloud solutions may be a powerful
way to reduce the need to understand the complexity behind digital
129
infrastructures, thus allowing executives to focus on the business side

instead of on digital infrastructures. The company just plug into the cloud
infrastructure, without worrying about how the hardware and software
layers are working. In addition, this increases the chances of having access
to updated and innovative applications, as there are economies of scale for
the cloud provider to keep the investments in new technologies, to keep
itself competitive in relation to other providers.
The same from the perspective of risks. Managing risks against digital
infrastructures is becoming increasingly more difficult. Using cloud
computing solutions means the service provider is responsible for the
security of part of the business. However, these both aspects of managing
complexity and risks can also be otherwise, if the company either is not
aware of the problems involving cloud services themselves, or if the
company chooses providers which do not have expertise in security. These
caveats are discussed in the next section on challenges.
Activity 11.4: Companies exploiting the advantages of cloud computing

Research online provides examples of companies who have got the benefits of using
cloud computing. Link the cited benefits to the list above. Try to find further advantages of
using cloud computing.
11.2.5 Challenges and risks of cloud computing

In this section we summarise the main challenges and risks faced by those
who use cloud computing services. Organisations should better prepare
for the challenges and risks, in such a way that they will be able to get
the planned benefits. In the next paragraphs, we explain some of these
challenges and risks, from technical and contractual perspectives, focusing
on the business use of cloud computing.
Data security and privacy: Any organisation is responsible for data
security and the keeping the privacy of their customers, business partners
and employers protected. Cloud services, as any other computing service,
is at the risk of attack by hackers. Although the cloud service is responsible
for keeping the security of its own servers, the fact is that when a data
breach happens, people are going against the companies they have trusted
the information with. There is thus a huge risk of reputation damage.
Naturally, this risk is not different from the one the company also has
when managing data internally. Most of the times, if an organisation
chooses good service providers, the cloud solution can have the same or
better security (see details in previous section).
Compliance with regulations and laws: The management of data is
protected by regulations and laws. These laws vary in accordance with the
country and international agreements. For instance, the Data Protection Act
in the United Kingdom enforces data sovereignty, protecting defined sets of
data. The Patriotic Act is the relevant legal framework for data protection
in the USA. The European Union’s Data Privacy Directive is valid in Europe.
Thus, before deciding to use cloud services, an organisation needs to
understand the regulations and laws, in order to be sure that storing and
managing data in the cloud is not breaching the relevant legislation. One
important aspect to check is the auditability of the services, to be sure that
the service provider can prove that they respect the legislation, as agreed in
the contract. If specific sets of data cannot be moved offshore, for instance,
the cloud provider needs to keep the data in servers inside the country and
needs to be able to prove this procedure.
130
Risk of lock-in situations: A key question when choosing the cloud

solution is to define the data portability. In other words, if a customer
decides to change the cloud provider, or to move the data to be managed
internally again, the data should be easy to transfer to the new dataset,
in open standard format. If a customer does not have the means of
transferring its data or of accessing the data through open standard
formats, it is a situation of lock-in: one cannot leave the service providers
without losing the data which is there. One needs to read the contracts
carefully to be sure they are not locking the company into a bad contract.
Similarly, a lock-in situation may happen because the service provider
does not allow interoperability between its service and others. In this case
the company only has the chance of using the services of one provider,
even when others could offer a better combination of options. Thus, the
customer should also check whether the cloud provider offers easy access
to the data, favouring interoperability.
Credibility of suppliers: Before contracting cloud services, an
organisation should research the credibility and reputation of the service
providers. The fact is that many times there is not much information on
suppliers and services, making an informed decision more difficult. Before
contracting services, the customer should clarify fundamental questions
such as the reliability in the provision of services (outages), the methods
used to measure the use of the services (metering), the capacity of the
supplier in investing in technology to keep the quality of the services and
the availability of customer support. Unclear pricing methods is an area of
conflict between customers and suppliers.
Looking for certified providers (for instance, checking those who have ISO
certification in critical areas) may be a way of verifying the reputation
of suppliers. It is necessary to check the Service Level Agreement (SLA),
avoiding contracts that are unclear. The imposition of standard SLA
may create problems for companies which have specific needs. This is a
particular problem when service providers may change the standard SLA
without previous agreement with the clients. Particular attention should be
given to contracts which are renewed automatically.
Also be aware that cloud providers may discontinue their services. This
situation would have serious consequences for customers in terms of
business continuity if there would not be a contingency plan and if the
provider does not inform them of the change within a reasonable time for
allowing a smooth transition.
Poor business case: There are two main problems with the business
case when planning cloud solutions. The first is at the strategic level,
when the company has not been able to align the use of cloud services
with the business objectives. This happens due to lack of understanding of
the cloud options or of the business, or both. In some cases, the company
executives are not interested in building this alignment as they wish to
keep control of the digital infrastructure budget, avoiding shifting power
to other organisations.
A second problem happens when the company does not calculate properly
the costs and benefits of contracting cloud services. This may be due
to lack of information about the services or for lack of understanding
of one’s own business. In this situation, it is not possible to calculate
the return on investment (ROI). In addition, the customer may face
obstacles to getting the benefits because of technical problems, such as
poor network connection. Cloud services are not going to work well if
the communication networks are not reliable and fast. Other technical
131
problems jeopardising the business case may be lack of understanding of

communication standards and interoperability between the cloud provider
and other systems and the need to have skills and capabilities above the
current level.
Activity 11.5: Companies facing the challenges and risks of cloud computing
Research online examples of companies who have faced problems for using cloud
computing. Link the cited problems to the challenges and risks listed above. Try to find
further risks of using cloud computing.

This chapter developed the concept of cloud computing, presenting a
formal definition for the term and explaining the technical underpinnings
of this definition. The chapter presented some historical facts about the
way cloud computing has evolved and a typology of cloud services and
deployment models. The chapter also demonstrated how cloud computing
is dependent on the innovation of digital infrastructures and an enabler
of business innovation. Finally, the chapter presented the key advantages,
challenges and risks of cloud services, with some recommendations on
how to get more benefits from this digital infrastructure.

• describe the concept of cloud computing and the underlying
assumptions behind the services
• compare the advantages and challenges of cloud computing to inform
decision making
• explain the connection between cloud computing and digital
infrastructures
• applying basic cloud computing to business innovation.

1. Define cloud services and explain the key technical aspects which
made the cloud infrastructure possible.
2. Critically evaluate the key advantages, challenges and risks of cloud
computing solutions.
3. Explain how cloud computing depend on the innovation of other
digital infrastructures and how cloud services have fostered innovation
in other business.
132
Chapter 12: Mobile technology as an innovation in infrastructure
Chapter 12: Mobile technology as an

Chapter outline
This chapter aims to explain the emergence and impact of mobile devices
as digital infrastructure. It discusses how computing has moved from
desks to the hands of users around the world and how it has changed
the way organisations and individuals understand computing. It brings
some technical elements to understand the emergence of mobile digital
infrastructures and it focuses on business and sociological elements to
understand the impact of mobile devices on economies and societies.
12.1 Introduction
One of the key characteristics of our contemporary societies is the presence
of mobile technology in our hands. From very rich to very poor countries,
the mobile phone is there, supporting communication and interaction. For
some, it is the device of choice for doing business and work on the road.
For others, it is the best way of keeping contact with family and friends,
sharing nice memories. Some would spend lots of time playing games in
the devices. Others would see an opportunity for preparing for an exam
using e-learning interfaces. All the many functionalities that are available
today through mobile technologies have become possible with the
development of smart interfaces and compact computing for processing
large volumes of data in small devices.
The more capacity the mobile device has to process data locally and
the more speed it has to transfer data, the more the operating system
evolves to support more services and the more developers are interested
in providing apps to run on the top of the operating system. Through this
path, mobiles have evolved from being just pagers (to send short texts
only), to become phones for conversations and later devices with huge
computing capacities, with GPS, internet browser, broadband connections.
Indeed, today mobile operating systems have evolved to become platforms
for all sorts of apps, which are specialised software applications which
run on mobile phones and tablets (Chapter 17 gives details on the
conceptualisation of platforms).
Thinking from the perspective of the hourglass architecture (Chapter 6),
there are many elements embedded in each layer of the current mobile
infrastructure. The final result we have in our hands is the combination
of hardware, software and a collection of services, which are supported
by other layers of hardware, software and services which are behind the
local infrastructure of mobile cell towers and corporate infrastructures
which manage the services offered to support the whole mobile industry.
All the parts of this complex infrastructure of layers upon layers rely on
communications standards which connect each element to each other,
allowing each part to develop with a degree of independency from the
whole systems, at the same time that the development of some elements
enable other advancements in other parts of the infrastructure. Every
time a standard is not well developed or not agreed across the many
stakeholders responsible for the whole system, parts of the functionality of
the mobile technology infrastructure is jeopardised.
133
As discussed in previous chapters, the digital infrastructures also

support each other to a great extent. There is a self-reinforcing feedback
mechanism through the interaction of mobile technologies with other
digital infrastructures (the internet, the WWW, cloud computing and social
media), which together attract more people to make mobile devices the
main channel to access the internet, social media and the services provided
by apps. The reason is simple: there is no value in a device without content,
in the same way that the value of content is smaller when the access to
it is more difficult. The easier the access to the internet, the more people
discover the value of internet content in their daily lives. Even better, the
mobile device is also equipped to provide communication and interaction
with friends, family and business contacts.
Considering the objective of this subject guide, this chapter focuses on the
application of mobile technology digital infrastructures. The chapter does
not provide a detailed account of the technology itself. Rather, the focus is
on the impact of mobile technology on businesses and beyond. Also, in this
chapter we take into consideration that mobile platforms, such as Google
and Apple, are elements of the layers of mobile infrastructures. Later in
Chapter 17 we discuss more about platforms, but in this chapter we need
to understand that technology, software and services come together to
make mobile infrastructures relevant. It is difficult to understand the value
of mobile infrastructures without understanding the services which are
associated with it.

• Introduce key ideas on innovation of mobile technology digital
infrastructures.
• Discuss the impact of mobile technology on communication.
• Present key impacts of mobile technologies on business environments.
• Explain how mobile technologies have enabled the app services.

• describe the impact of mobile technologies on organisations,
individuals and societies
• explain the role of app services as layers of the mobile technology
infrastructures
• research and identify apps for particular objectives in mobile operating
systems
• discuss some of the risks of mobile technology.

Laudon, K.C. and J.P. Laudon Management information systems – managing the
digital firm. (Harlow, England: Pearson Education, 2021) 17th edition
[ISBN 9781292417752]. Read all the parts of the book related to cloud
computing. See in particular Chapter 5.
Mobile technology: https://en.wikipedia.org/wiki/Mobile_technology

Bezerra, J. et al. ‘The mobile revolution: how mobile technologies drive a
trillion-dollar impact’ BCG – Boston Consulting Group (2015). https://www.
bcg.com/publications/2015/telecommunications-technology-industries-the-
mobile-revolution
134
Standage, T. Writing on the wall. (London: Bloomsbury Publishing, 2013)

[ISBN 9781408842065].

Bezerra, J. et al. (2015). The mobile revolution: How mobile technologies
drive a trillion-dollar impact. BCG – Boston Consulting Group. January
2015. https://www.bcgperspectives.com/Images/The_Mobile_Revolution_
Jan_2015_tcm80-180510.pdf
Carr, N. The shallows: what the internet is doing to our brains. (New York:
W.W. Norton & Company, 2011) [ISBN 9780393339758].
comScore The 2016 US Mobile App Report (2016). Whitepaper https://www.
comscore.com/Insights/Presentations-and-Whitepapers/2016/The-2016-
US-Mobile-App-Report
Meola, A. ‘The rise of Mcommerce: mobile ecommerce shopping stats and
trends in 2022’ Business Insider https://www.businessinsider.com/mobile-
commerce-shopping-trends-stats?r=US&IR=T
Meyer, K. ‘Millennials as digital natives: myths and realities’, Nielsen Norman
Group (2016). https://www.nngroup.com/articles/millennials-digital-
natives/
Morozov, E. The net delusion: How not to liberate the world. (London: Penguin
Books, 2012) [ISBN 9780141049571].
‘PayPal mobile research 2014/2015: global snapshot’ (2015) https://www.
paypalobjects.com/webstatic/en_US/mktg/pages/stories/pdf/paypal_
mobile_global_snapshot_2015_2.pdf
Standage, T. Writing on the wall. (London: Bloomsbury Publishing, 2013)
[ISBN 9781408842065].
Statista (2021). Number of apps available in leading app stores as of 1st
Quarter 2021. https://www.statista.com/statistics/276623/number-of-
apps-available-in-leading-app-stores/
https://www.statista.com/statistics/330695/number-of-smartphone-users-
worldwide/
https://www.statista.com/statistics/539395/smartphone-penetration-
worldwide-by-country/
https://www.statista.com/statistics/1133365/mobile-penetration-sub-saharan-
africa/
https://www.statista.com/statistics/558610/number-of-mobile-internet-user-
in-india/
https://www.statista.com/statistics/806336/mobile-retail-commerce-share-
worldwide/
https://www.statista.com/statistics/806323/mobile-retail-commerce-revenue-
worldwide/
https://www.statista.com/statistics/276623/number-of-apps-available-in-
leading-app-stores/
Wikipedia (on WeChat): https://en.wikipedia.org/wiki/WeChat

This chapter develops the discussion on mobile technology digital
infrastructure, focusing on the impact of these devices in businesses and
beyond. Special emphasis is given in the impact of mobile technology in
communication patterns and behaviour. The chapter brings information on
mobile technology innovation and how this has enabled further innovation
in the business side. The chapter applies the concept of hourglass
architecture throughout the analysis and link mobile technologies with
other digital infrastructures.
135

12.2.1 The pervasiveness of mobile technology
The innovation of mobile technologies in the last two decades has changed
the perception of users about the benefits of these devices. On the one hand,
there were great innovation in the technology itself. Mobile phones and
personal computing have converged to one device (smartphones), which
have gained better interfaces (touching screen). The data transmission has
become much faster (wireless broadband, such as the 4G network) and the
spread of mobile cell towers have supported the expansion to the whole
network. On the other hand, there was great innovation in the services
enabled by mobile technology. Operating systems (such as iOS and Android)
have become more robust and flexible, offering open doors (via APIs –
Application Programming Interfaces) for other developers to create apps and
plug in their pieces of software into the mobile infrastructure. These apps
have then been made available through mobile commerce services, such as
Apple’s App Store and Google Play (see more on digital platforms in Chapter
17). Payment systems have been added to the system, through credit carts
and electronic solutions such as PayPal. Public spaces and some businesses
have added a layer offering free WiFi to local users.
Considering the flexibility and functionality of mobile technology, it is not
a surprise that the sales of smartphones have passed the sales of computer
desktops and the number of users of smartphones is expected to increase
from 6.7 billion to 7.7 billion people from 2022 to 2027 (estimated by
Statista, 2022). In the USA, for instance, people buy more mobile devices
than desktop computers for personal use, spending more time interacting
with mobiles than with desktops (comScore, 2016). In mature markets
such as the USA, the United Kingdom and Germany, the penetration of
smartphones is around 80% (Statista, 2022). And consumers expect more
innovation, with faster networks, more coverage, longer batteries and
other improvements to guarantee that the mobile devices can deliver better
services at any time, at any place.
The access to wireless broadband services is a fundamental reason for
the increasing use of smartphones and other mobile devices. Access to
communication infrastructure has been a huge obstacle for developing
countries to use more computing technologies. With wireless broadband, it
has become easier to overcome some of the barriers to spread computing and
communication infrastructures also among developing countries. The change
is such that in Sub-Saharan Africa, the penetration of mobile technology
covered 46% of the population in 2020, a number which is expected to
grow to 50% in 2025 (Statista, 2022). In India, most internet users access
the network through mobile devices. In 2020, the internet population in
India was of 749 million users, being that 744 million people accessed the
internet through mobile devices (Statista, 2022). This shows that developing
countries are overcoming their disadvantages in terms of communication and
computing infrastructures through the advances in mobile technologies.
Activity 12.1: Mobile technology penetration in different countries

Mobile technologies are quickly spreading around the work. However, there is still a big
gap between some rich countries (in which people can afford to have two or more mobile
devices) and poorest countries (in which people need to share the same device with
family and friends). Research the level of penetration of mobile technology in different
countries. Compare and discuss the consequences of these differences. What does it
mean to be a poor person without access to the internet? What sorts of educational and
professional opportunities are those without internet access missing?
136
Activity 12.2: Understanding mobile technology in Africa

Why has Kenya lead the world in mobile money? Read about the M-Pesa mobile phone-
based payment system in Kenya and other countries in Wikipedia.
The potential of mobile healthcare in Africa: Read about the potential for mobile phone-
based healthcare – m-health – as part of social development in Kenya.
https://www.slideshare.net/economistintelligenceunit/the-future-of-healthcare-in-
africa-41501807
Business benefits
Businesses have strongly benefited from mobile technology infrastructures
in the last decades, favoured by the reduction in costs of communication
and increase in the speed of access to data. Research prepared by Boston
Consulting Group (BCG) (Bezerra et al., 2015) shows that the cost for
using mobile devices for accessing data has decrease more than 90% from
2005 to 2013. This has happened mainly because the new generations of
mobile network technologies are much faster, with a massive migration
of users from obsolete networks to new ones. To give an example, the 4G
networks are 12,000 times faster than the 2G networks. Consumers accept
paying more for getting more efficient access to content in their mobile
devices.
In a broad sense, mobile technology brings interactivity and access to
information, enabling businesses to offer new products and services and
better customer engagement and to change operational processes to gain
productivity. For instance, employees may have access at any place to the
customer relationship management system and give a better advice to
the client. Customers may search for prices and product descriptions and
pay for their purchase in their mobile devices, or they may even use the
mobile phone for making payments (for instance, using NFC – Near Field
Communication). Employees may work from any place, saving commuting
time and giving better support for customers in the place of their choice.
Collaborative technologies available through mobiles give easy access to
experts. Integration of calendars and contacts in the mobile makes easier
for professionals to manage their diaries and find the people they need at
the correct moment. Localisation services help to provide better services
when your customer is at the door, such as offering a free coffee to a
regular customer and allowing your sales professionals to use GPS services
to quickly find the directions to the client.
Researching the impact of these technologies on SME businesses
(small and medium companies), BCG has concluded that there is a
substantial benefit for the companies which have led the adoption of
mobile technology infrastructures. Considering a survey with around
3,500 SMEs in the USA, Germany, South Korea, Brazil, China and India,
the research shows that these companies have benefited in terms of
reaching more customers, responding to the expectations of customers
in terms of availability of business in mobile devices and growing faster
than competitors. These leaders claim that mobile technologies have
allowed the company to pursue new business opportunities and to
change the business processes. Finally, they claim that mobile technology
has improved their access to real time information, increasing business
productivity and capacity of innovation.
137
Activity 12.3: Mobile payment technology

SMEs have been using mobile technologies to innovate their own business models and
aggregate value to their customer. One interesting technology which has changed the
operations of some SMEs is the Block mobile payment (formerly known as Square).
Through a reader connected to their mobiles, SME retailers and consultants may receive
credited card payment for their services and products. Check online on the link below
about the service and research online for companies which have benefited from using
Square. You can find other similar products, such as iZettle.
https://en.wikipedia.org/wiki/Block,_Inc.
https://en.wikipedia.org/wiki/IZettle
The customer goes mobile

From the customer perspective, mobile technologies have become a
relevant channel for purchase and payment. Worldwide, the mobile
retail commerce sales represented 73% of the total retail e-commerce
sales in 2021 (Statista, 2021) (in monetary value). The volume of
sales through mobile phones has increased substantially because of the
Covid-19 pandemic: from US$ 2.3 trillion in 2019 to US$ 3.6 trillion
in 2021 (Statista, 2022). These numbers confirm the relevance of
mobile technologies in supporting e-commerce. There are implications
for businesses, as they need to guarantee their presence in the mobile
platforms (Android and iOS, for instance), to reach customers.
Although the volume of sales is impressive, many users are still
conservative in adopting m-commerce. In a survey in 2015, PayPal found
that only 33% of mobile users have purchased something through their
devices (worldwide). Another survey showed that, in 2022, only around
40% of e-commerce sales were made through mobile devices in the USA
(Business Insider, 2022). It shows that there is still plenty of space for
more users to adopt m-commerce, which is a market opportunity for
companies (or threat, for those businesses which decide to ignore the
trend). PayPal’s survey showed that the first step for consumers is to use
the mobile device for searching for information on products and suppliers
and checking the reviews given by other consumers. Responsive design
should be used, to allow the company’s website to identify the device of
the customer and adapt the content to the correct device. The purchase
comes later when the consumer is more confident on the security of the
mobile device and payment systems. Consumers also give preference
for using an app to make the purchase, instead of using the normal
browser (47% prefer the app, 27% prefer the browser and 25% can use
both), another business opportunity for many businesses which want to
differentiate themselves from competitors. Apps are easier to use and can
store the preferences of the customers and facilitate the payment.
Bring your own device (BYOD)

From the employee perspective, personal access to advanced mobile
technology has changed their expectations on which technology the
companies should adopt. In some companies, there is a conflict between
the expectations of employees and the investments done by organisation.
This situation is very different from the time in which people would expect
to find the most updated technology in the workplace, not in their own
pockets. For overcoming the challenge, some companies have introduced a
policy known as ‘bring your own device’ (BYOD), which allows employees
to bring their own mobile devices for working in the business and for
138
accessing the company’s data and information from any place. Some
consultants call this phenomenon ‘IT consumerisation’.
Instead of fighting with employees, many companies have accepted
the BYOD movement, seeing also an opportunity of benefiting from an
investment which is done by the employees. Advocates of the model argue
that BYOD increases productivity, as employees invest their own time to
learn how to use and maintain the device and are also more motivated
for having chosen their own working tools. On the other hand, there are
serious security risks: when not properly protected, these devices can be
lost or hacked, giving criminals an entry door to the company’s data and
businesses. For this reason, companies which adopt the BYOD approach
need to give employees support for the use of the technology and need
to have special measures to protect the security of its assets. Encryption
of devices and clear policies for accessing and storing data are among
the measures which help the company to protect its information when
employees use their own devices.
Activity 12.4: The risks of BYOD

The phenomenon of BYOD has brought other security risks for organisations and
employees. Research online the key risks for organisations to adopt the BYOD solution.
Discuss the solutions to mitigate these risks.
12.2.2 Changing communication patterns and behaviour

Mobile technology infrastructures have brought a huge change for human
communication. For most of our history, we have either communicated
in face-to-face encounters, or we have made ourselves available at a
particular address for defined messages to reach us (see more on the
history of communication in Standage, 2013).
Let’s think about letters. Once upon a time, letters were the way to
communicate with people across large distances, when one could not
meet the other in person. For one to receive a letter, it is necessary to
have a defined address, a place a person either is expected to be or is
expected to come regularly to get the messages. In medieval times, for
people to know the news, they either would go to a central place in a city
or village, in which news were spreading from other cities and villages
through travellers, or they would wait by the chance of having a person
coming to bring the news to them in their own place. Even when the
communication channels have evolved to telegraphs, changing the speed
of communication, still the receiver of a message should be in a particular
address. Thus, place has been very important for most of human history
when we think about the way we communicate.
Computing has changed this pattern. The first change is about emails. For
the first time, the message would be available in a virtual space (indeed
a server) and people could access this email server using any computer.
However, the computer at that time also would be in a fixed address.
Computers to access networks would need to have access to a physical
infrastructure which were available only in fixed addresses, such as offices
and houses. One could access the email from different places, but in each
of these places, there will be a prepared infrastructure of communication
cables to allow the email to be read and answered. A huge step in terms of
flexibility – one did not need to be in a defined place to receive a message
– but still with limitations.
139
Mobile technology has liberated us from place, in a large degree. We can

now communicate from any place using our mobile devices (laptops,
smartphones, tablets and wearables). There are still limitations, as this
mobile communication depends on the installed local infrastructure
which supports the mobile devices. Many times, your mobile device does
not have a good connection, as the mobile cell towers do not have the
same level of quality in all places. If you have a smartphone based on 4G
networking, for instance, you can find yourself without communication
in places the network infrastructure has not advanced yet. You can travel
to places in which the local service infrastructure does not support the
communication with your device (poor roaming services). You can also opt
out for using your device fully when travelling abroad for reasons of costs,
as the roaming services are more expensive than the regular services,
particularly for the transmission of data. In spite of the limitations, we
have never had such a flexibility in our communication before mobile
technologies. And we can foresee that most of these current limitations are
going to be resolved in the future.
Implications of new communication patterns

The changes in communication patterns have implications. For instance,
exactly because now we are available independently of place, we can
also observe changes in the expectations about when employees should
be available for working. This is not a trivial question. In the past, people
would go to the office for working and would go home for having their
own life. There was a clearer separation between working and personal
life. Nowadays, people are available 24/7 in their mobile devices. Should a
professional be available to answer the emails in the weekend? And during
the vacations? In spite of the legislation on these matters, which mostly
protect the workers from overworking or working without payment, there
are social norms which in practice override the legislation. The change in
the communication patterns has thus also changed the working patterns,
with more people not being to differentiate the working time from the
personal life. The boundaries between work and personal life have become
even more blurred after people have been forced to work from home
during the Covid-19 pandemic.
On the other hand, we also need to define when the moment for
communication related to our own life is. Today more people use part of
their work time to manage their private communication using their mobile
devices. Some companies understand that the employee should not use
working time to manage their private businesses. However, in practice,
it is mainly impossible to control what people do (at least in democratic
countries, which have legislation protecting citizens and employees).
A solution for companies is to change the focus of measurement from
behaviour to outputs: as it is not possible (or legal) to ensure a particular
behaviour is adopted, checking the quality and quantity of outputs become
a better way of managing people. This means other forms of management
are emerging, with smaller number of managers (less hierarchical layers)
and more digital control of outputs (for instance, how many customers have
their problem sorted out in less than five minutes in a call centre). To be
clear, this phenomenon does not happen only because of mobile technology;
rather, it is a result of the increasing use of information technology in
general. However, mobile technology has reinforced this trend.
Changes in the communication patterns have also affected other aspects
of our behaviour in society. It does not take much time to see this in
the streets. You enter any public transportation system or any coffee
140
shop or restaurant and observe how many people are interacting with
their mobiles at any given moment, even when they are in face-to-face
conversation with other people. Indeed, one can probably see this at home
too. Once people form an opinion on the combined benefits of mobile
technology and start using this technology everywhere, it becomes more
difficult to imagine one’s life without this interface.
The BCG’s research (Bezerra et al., 2015) has shown that more than
50% of mobile users would give up of going out for dinner, having a
pet, or going on vacation it this sacrifice would be necessary for keeping
their mobile devices. More than 50% would also give up a day off work
per week in exchange for mobile connection. Happily, we can keep
all. However, this research shows that mobile technology has become
a fundamental part of our lives, information which is confirmed by
many other research reports. One can question, however, whether this
happens because of actual benefits or because of perceived benefits, or
even because of social norms: as everyone is online 24/7, as everyone is
communicating through social media, no one wants to be left behind.
Activity 12.5: Being online 24/7

Research the legislation of your country to understand if or how the law protects
employees from being available 24/7 to answer emails. Also research what professionals
and companies think about the topic in your country. What are your conclusions? In your
country, are the social norms aligned with the legislation on this matter?
Privacy and surveillance

Our understanding of privacy and etiquette is also affected by the changes
in communication patterns brought by mobile devices. As we can talk
to people anywhere, the boundaries of what can be said in public have
become blurred, as have the boundaries of where we could use our mobile
devices. In any big city around the world one can stop at a busy street
and try to listen to others’ private conversation. People act as if they were
talking inside a transparent bubble when they are using their mobiles.
But there is no bubble. Very interesting topics may be listened to in many
public places these days.
Let’s think about a regular lecture in a regular university in a western
country. One can bet that there will be a reasonable number of students
checking their social media accounts during the lecture. The use of mobile
devices is not perceived to be offensive or distracting. Or is it? These are
areas in which new social norms are to be discussed and agreed. The myth
of multitasking has been already challenged by research (see the debate
at Meyer, 2016). If we are using our mobiles at the same time as we are
sorting out other problems, chances are that we are not doing any of the
tasks properly. Some authors argue that at the cognitive level we are losing
our attention span and ability to concentrate (Carr, 2011). These studies
are generic about internet use as a whole, but mobile devices are the
culprit that makes the internet and social media available at any time. See
more on this discussion in Chapter 13.
Finally, mobile technology brings the risk of more surveillance (some
would say the opportunity of having more information about individuals).
Having the mobile with us all the time means that it is always possible
to know where we are. At least the mobile service provider knows and
most of the times many other developers of apps which require to know
our localisation. Specifically, companies can install apps in employees’
141
mobiles to track their whereabouts, or the whereabouts of their cars.

Unfortunately, this has also happened in the private sphere, when partners
track each other. It is also possible to know what we have written to others
at any time. If necessary, a court order may require the service provider to
pass the communication of a person for further investigation.
Two examples may illustrate how surveillance can be used. First, in
August 2011, there were a sequence of riots in London. People organised
themselves using mobile phones to message each other. The organisation of
the attacks was very decentralised and difficult to control. In the following
months, British courts requested details on the mobile communication of
thousands of people, suspected of being involved in the riots and many of
them have been condemned based on their mobile communication. Second,
in 2013 and 2014, Ukrainians decided to challenge the government of the
president Viktor Yanukovych, occupying the Maidan square in Kyiv. The
government used mobile signals to track the identity of those who were in
the protest. See more on surveillance in Morozov (2012).
12.2.3 App services

One key change brought by smart mobile interfaces is the emergence
of app services. Part of the software layer of the mobile technology
infrastructures, the apps are the interface to a myriad of specialised
services, from social networking and calendar to gaming and healthcare.
It is difficult to estimate with precision the number of apps around the
world. The Statista estimates that there are 2.2 million apps running in
the iOS operating system (sold by Apple’s App Store), 3.5 million apps
running in the Android operating system (sold by Google Play) and almost
700 thousand apps in the Windows operating system (Windows Store) (in
2021). WeChat, which works as a super-app with sub-programmes inside,
has the equivalent of more than 1 million apps (or mini programmes,
depending on how you name it). The data for running the apps are part
in the mobile device and part in the cloud services, thus the quality of
the broadband connection is an important aspect to keep the high-quality
experience the final user is expecting.
App services have multiplied the value of mobile technologies. Each user
customises their mobile device through the choice of apps and through
the way the apps are arranged to achieve a particular configuration and
deliver aimed outputs. This capacity of combining apps from a poll of
millions of options and of deciding how each app is going to work in one’s
mobile, mean that each mobile device is different from others. This is the
most flexible computing technology we have ever had. The app layer has
brought to the hands of users the capacity of innovating: in recombining
apps to individual needs, each user has the potential to create their own
solution to a particular problem.
App services enable other innovations such as the personalisation of
services. The more the mobile operating system knows about the chosen
apps, the easier it is to recommend other apps or to understand the profile
of the user and to use this information in other interfaces. For instance, if
a user has lots of games on their mobile phone, it may be a good idea to
add more advertisement on games next time the user is browsing apps.
Going beyond, as the key mobile operating systems (such as Google, Apple
and Microsoft) integrate the mobile experience with the other devices used
by the individual, having downloaded a mobile phone app on learning to
cook may bring you advertisements on kitchen utensils and devices when
using your computer for search.
142
In Chapter 14, we come back to this topic of app services, discussing more
on the sort of functionalities they offer.
Chapter 12.6: The app services

The key mobile operating systems offer millions of apps. It may be difficult to find
interesting apps within a large range of options. Research online how apps are ranked by
users and how rankings help users to find the apps they need. What are the other means
a user may use to find an app?

Focusing on mobile technology digital infrastructure, this chapter brought
key information on the innovation related to these devices and the impact
of these innovations on human communications and businesses. The
chapter summarised some technical aspects and crunched a few numbers,
highlighting, however, the changes in business and society. The chapter
ended emphasising the relevance of app services for increasing the
value of mobile technologies and fostering innovation, a topic which is
developed further in Chapter 14.

• describe the impact of mobile technologies on organisations,
individuals and societies.
• explain the role of app services as layers of the mobile technology
infrastructures.
• research and identify apps for particular objectives in mobile operating
systems.
• discuss some of the risks of mobile technology.

1. Apply the hourglass architecture model to explain the interaction of
different layers in the mobile technology digital infrastructure.
2. Critically discuss the impact of mobile technology in human
communication.
3. Discuss how the diffusion of mobile technologies has affected
businesses.
143
Notes
144
Chapter 13: Social media as an innovation in infrastructure
Chapter 13: Social media as an

Chapter outline
This chapter aims to conceptualise the social media phenomenon and to
explain the empirical relevance of social technologies and services. The
objective is to provide some conceptual frameworks to allow students
to better understanding the emergence and relevance of social media
mainly from the perspective of organisations. The chapter explores some
applications of social technologies in business, with especial attention to
its deployment in collaboration, innovation and marketing.
13.1 Introduction
Social media is the generic name we use to conceptualise a category of
digital tools that allow individuals and organisations to create and share
information and knowledge, to communicate and to collaborate with
each other through computing infrastructures. These communication and
collaboration efforts may be in closed groups (such as on Facebook and
LinkedIn), or they may be in public spaces in which anyone can participate
(such as Twitter and some blog services). Some people prefer to use the
term social technologies to refer to these tools, including in this case the
proprietary tools offered to companies to foster collaboration among
employees and contributors (such as Microsoft’s SharePoint and Teams
and IBM Connections).
The fundamental aspect of this definition is the capacity for interacting
through social media, using Web 2.0 applications (which allow people to
publish their own content and comment and rate the content of others).
In this direction, social media offer a tool for people to generate their own
content and to build their own social networks. The value of social media
infrastructures emerges from the interplay of an enabling tool (which
allow people to create profiles, write content, publish photos and videos
and connect to others) and millions of individuals who believe there is
value in interacting online with other people. There would not be social
media if the user would not be interested in generating content for sharing
in these interfaces, and in giving a great amount of time to interact with
the content of others too.
Social media started, nonetheless, a long time before the facilities of
Web 2.0. The first interfaces of social media were computers connected
to private networks, which would allow users to access discussion lists.
These discussion lists would be hosted by organisations, which would have
dedicated servers to keep track of the email conversation. The idea behind
the creation of discussion lists was that computing technology would
enable people to form communities mediated by digital technologies.
These communities would be able to share content, but also would also
allow people to give emotional support to each other. At that time though,
the concept of cloud services was not fully developed. Then individual
organisations would need to build their own digital infrastructure to host
the discussion lists in their servers. Today, companies such as Facebook,
LinkedIn and WeChat offer the services in the cloud, thus nobody needs to
concern about investing in their own digital infrastructure.
145
So, the idea of building communities was present in the very beginning of
the first conceptions of social media, although at that time nobody called
these tools ‘social media’. Thus, this chapter focuses on the community
roots of social media and how this affects the use of these tools for
collaboration, innovation and marketing.

• Provide further conceptualisation of social media infrastructures.
• Explain the relevance of social media for fostering collaboration and
innovation.
• Introduce some ideas on how social media is used for marketing.
• Highlight some risks and challenges of social media tools.

• explain how social media foster collaboration within organisations and
beyond
• describe the link between the idea of collaboration and the process of
innovation
• discuss key risks and challenges associated with social media tools
• apply some frameworks to understand the use of social media by
organisations.

Social media: https://en.wikipedia.org/wiki/Social_media
Enterprise 2.0 (as defined by Andrew McAfee): https://en.wikipedia.org/wiki/
Enterprise_2.0
Animation presenting the summary of the book, Where good ideas come from,
by Steven Johnson: https://www.youtube.com/watch?v=NugRZGDbPFU

Kietzmann, J.H., K. Hermkens, I.P. McCarthy and B.S. Silvestre, ‘Social media?
Get serious! Understanding the functional building blocks of social media’,
Business Horizons 54(3) 2011, pp.241–51. This video is a summary
of the article above, with interesting animation: www.youtube.com/
watch?v=BCjCbqYs_j0&feature=youtu.be
McAfee, A. Enterprise 2.0: new collaborative tools for your organization’s toughest
challenges. (Boston: Harvard Business Press, 2009) [ISBN 9781422125878].
In this video Andrew McAfee presents key ideas on Enterprise 2.0:
www.youtube.com/watch?v=6xKSJfQh89k
The Economist Intelligence Unit ‘Networks for thinking: developing ideas and
forming opinions in the digital age’ (2011)

Bolstrom, N. Superintelligence: path, dangers and strategies. (Oxford: Oxford
University Press, 2014) [ISBN 9780199678112].
Brown, J.S. and P. Duguid ‘Organizing knowledge’, California Management
Review 40(3) 1998, pp.90–111.
Carr, N. The shallows: what the internet is doing to our brains. (New York: W.W.
Norton & Company, 2011) [ISBN 9780393339758].
Chui, M., J. Manyika, J. Bughin, R. Dobbs, C. Roxburgh, H. Sarrazin, G. Sands
and M. Westergren ‘The social economy: unlocking value and productivity
through social technologies’ McKinsey Global Institute (2012).
146
Deiser, R. and S. Newton ‘Six social-media skills every leader needs’ McKinsey
(2013)
Johnson, S. Where good ideas come from: the natural history of innovation.
(New York: Allen Lane, 2010) [ISBN 9781846140518].
Kietzmann, J.H., K. Hermkens, I.P. McCarthy and B.S. Silvestre ‘Social media?
Get serious! Understanding the functional building blocks of social media’,
Business Horizons 54(3) 2011, pp.241–51.
McAfee, A. Enterprise 2.0: new collaborative tools for your organization’s toughest
challenges. (Boston: Harvard Business Press, 2009) [ISBN 9781422125878].
Smith, N., R. Wollan and C. Zhou The social media management handbook:
everything you need to know to get social media working in your business.
(Hoboken, New Jersey: John Wiley & Sons, 2011) [ISBN 9780470651247].
Statista websites: http://www.statista.com/ and http://www.statisticbrain.
com/youtube-statistics/
Tuten, T. and M. Solomon Social media marketing. (Harlow: Pearson, 2013)
[ISBN 9781292023533].

Drawing upon academic and practitioner research, this chapter introduces
more details on social media as digital infrastructures (complementing
Chapter 3), from the perspective of organisations. The chapter pays
particular attention to the discussion on how social media may foster
collaboration and innovation and how social media is used for marketing.
Considering the limitations of this subject guide, only a few frameworks
can be discussed in each of these topics. Finally, the chapter highlights a
few risks and challenges brought by social media interactions. There is
much more to say on social media, also from the individual and societal
perspectives, but we need to limit the discussion considering the objectives
of this subject guide.

13.2.1 The relevance of social media for society and organisations
Never before in human history have so many people have been able to
communicate with each other using common channels. This is the biggest
change brought by social media. Think about more than 2.9 billion people
(in 2021) connected to the same communication interface, Facebook.
Then sum up the millions of users of other social media services such as
Twitter, LinkedIn, Instagram, Pinterest, WordPress, Tumblr, WhatsApp and
YouTube. Only YouTube has more than 2 billion unique visitors per month.
These numbers demonstrate the relevance of social media as digital
infrastructures.
Activity 13.1: Check new statistics on social media users

The numbers cited above are related to 2021. Research online more recent numbers
using the Statista website (see references above), Wikipedia and other sources. Check
which tools are more important in terms of number of users.
We cannot forget to say that the Chinese government has blocked many
of these social media services in China, in order to keep control over
the communication of their citizens. For this reason, we can find other
social media tools available in China, with similar functionalities, but all
controlled by the Chinese government. For instance, Chinese people use
the microblogs Sina Weibo and Tencent Weibo, the social network WeChat,
the instant message services QQ and the video sharing interface Youku. All
these tools have very similar functionalities to the social media channels
147
we focus on in this subject guide. We can see then that many more people
are using social media, even though we may not know about these tools.
WeChat alone has around 1.3 billion users (2021).
In the case of social media, size matters. The fact that people are using the
same social media service to communicate with each other speeds up the
process of communication and of getting new connections. Think about
your own experience. You meet a new person and would like to keep in
touch with them. Immediately you connect through social media tools, such
as Facebook, LinkedIn and WhatsApp. If you like the opinions of a person,
you just follow them on Twitter and check to see whether they have a blog.
In this way, you preserve this valuable connection for the future. The fact
that we are using similar communication channels leverage the capacity we
have of building and maintaining connections in the long term.
You also can get to know people through digital interactions. You may
belong to a discussion group about a topic and you may find other
interesting people with whom you would like to be connected, even if
only through digital intermediation. You see some friends of your friends
publishing interesting ideas and you connect to them, or follow them. You
choose with whom you would like to connect (and vice versa, others may
connect or not with you) and the terms of the interaction are to be agreed
between both, considering the limitations of the tools and the level of
privacy defined by the parts.
There are a great variety of social media tools and different authors
have classified these technologies in different ways. A comprehensive
classification has been proposed by McKinsey Global Institute (Chui et al.,
2012):
• social networks (enable connections through profiles)
• blogs/microblogs (focus on publishing content, although it also allows
connections)
• ratings and reviews (evaluation of products, services and opinions)
• social commerce (group purchasing and sharing opinions on products
and services)
• wikis (creation and storage of knowledge)
• discussion forums (spaces for debate and access to expertise)
• shared workspaces (co-creation of content and projects)
• crowdsourcing (harness collective knowledge or pool of resources)
• social gaming (game interfaces for playing with others)
• media and file sharing (sharing of content).
There are many tools which would be classified in one or more of these
categories.
In global research, McKinsey Global Institute has concluded that companies
are using social technologies for a variety of applications (Chui et al.,
2012). For instance, for product development, social media is an excellent
channel for getting insights from customers and even to engage customers
and contributors in the co-creation of products. For operations, social media
may help forecast a particular context and distribute work. For marketing
and sales, social media has become one of the main channels for advertising
and branding (see more in Chapter 14), in addition to bring customers’
insights, generating sales leads and favouring social commerce. Related to
marketing, it also allows a better customer service. For business support,
social media improve collaboration and communication within the firm
148
and beyond, enabling an easier system for matching expertise and skills to
defined tasks.
This report (Chui et al., 2012) concludes that a large majority of
companies benefit from using social media as channels for getting
quicker access to knowledge, from internal and external experts. On the
top of this, communication, travel and operational costs are reduced.
Overall, companies with good use of social media have seen increases
in productivity, reducing, for instance, the time spent on emails and
searching for information. However, this does not work for everyone and
in every situation (see section on risks and challenges in Chapter 14).
The holy grail of getting more benefits from innovation needs to be better
explored though.
Activity 13.2: The benefits of social media to companies

The paragraphs above have described how organisations benefit from social media. Read
the paper Chui et al. (2012) for a better understanding of this research. In particular,
find out which departments and which industries benefit most from using social media.
You can research other papers online to investigate the different uses of social media in
divers industries.
Focusing on this aspect of connections Andrew McAfee (2009) has defined

a useful concept about social media, which he calls Emergent Social
Software Platforms (ESSPs) (note that McAfee uses the word ‘platform’
in a slightly different way from how we define this concept in Chapter
17). First, social software refers to the fact that these tools allow people
to connect and collaborate online, forming digital communities. Second,
these tools are platforms because they offer a variety of features to be
combined and the results of the collective contributions are visible and
durable over time. Thirdly, they are emergent because the tools do not
define the way people are going to interact; rather, the social structures of
interaction are enacted by the people using the tools.
Social media tools do not impose defined structures of interaction. There
are limitations, but they aim to be more flexible in order to allow the user
to define the content and the form of the interaction. The flexibility also
comes from the capacity for organising the interaction through combining
different social media tools. We can see that each individual social media
tool enables particular outcomes and that a more powerful result comes
from the combination of different tools. Thus, the relevance of having
social media tools which can be combined for fostering open collaboration
and innovation.
McAfee argues that social media enable the emergence of the Enterprise
2.0: companies that use social technologies to foster new forms of
organising which are more collaborative in nature, in contrast with more
traditional, hierarchical forms of organising. Instead of seeing companies
as different silos of expertise which do not communicate to each other
easily, this perspective fosters the visibility of knowledgeable people and
the interaction as a way of creating and sharing knowledge. Indeed,
researchers argue these collaborative interfaces are to foster innovation.
We discuss this in the next section.
13.2.2 The emergence of collaborative interfaces

There are many arguments as to why companies should foster more
collaboration through social media tools. One perspective is that through
social networks a company can get access to people and databases they
could have not connected without these digital tools. McAfee developed
149
the concept of Enterprise 2.0 based upon the idea that social media
reinforces our weak and strong ties with people we already know and
simplifies the connection with people who we do not know yet. Access
to people is thus a key benefit of social media: a company may get easier
access to experts and creative customers, for instance.
An alternative view focuses on knowledge, emphasising that knowledge
emerges from the interaction among people. Taking this perspective, we
could see the benefit of using social media to enable the formation of
communities of experts. The pool of knowledge which emerges from such
a community is bigger than the parts: knowledge is recombined through
conversations and experts can rely on each other’s strengths to reach new
levels of understanding. In this direction, the concept of communities of
practice may be helpful here (see more details on Brown and Duguid,
1998). These communities are organised groups of people who share
common concerns and passions in a particular area of knowledge and are
willing to share their knowledge and concerns with others within these
groups (which may be restricted to the boundaries of an organisation
or be broader communities organised by digital interfaces). From this
perspective, social media is an interesting way of facilitating and fostering
the communication in communities of practice.
A third view of the relevance of social media collaboration comes from
authors who study innovation processes. Doing a historical analysis of
innovations for the last centuries, Steven Johnson (2010) brings a very
interesting view on the value of collaboration. He argues that what matters
for innovation is not only the number of nodes in a network, but the way
these nodes are connected and the quality of their interaction. In order to
have more innovation, people should be inside environments in which all
are exploring new boundaries. In this direction, the internet and the social
media are infrastructures which favour innovation, through improving the
connections and the quality of the interactions among experts.
In a further analysis, we can also discuss how social media may enable
humans to create new social structures for knowledge creation and
innovation. The idea behind this thought is that we can create spaces
for collective superintelligence, through new forms of organising and
linking individual human minds and artefacts (see more in Bolstrom,
2014). The idea is that superintelligence could emerge from the positive
network effect of connecting people and content in particular ways,
breaking complex problems in manageable equations, for instance. Digital
infrastructures in general provide means for large amounts of content
to be available and reachable – think about the internet and the WWW.
In particular, social media tools provide the means to connect people,
enabling news ways of organising collaboration. As communication
costs are reduced and social media tools provide details on the profile of
individuals, it is not difficult to foresee the emergence of algorithms to
foster collaboration and innovation. Today the focus of algorithms is on
marketing and searching, but the technology behind algorithms is evolving
to other areas. The more people are educated and connected through
social media infrastructures, the more we can expect developments in the
direction of fostering collective superintelligence.
Using an economic perspective to interpret the benefits of using social
media, digital collaborative technologies allow more direct interaction
between participants. This makes part of the coordination efforts
redundant, allowing the company to reduce the managerial levels in a
hierarchy. The speed of the process also could be increased, again reducing
150
costs and improving competitiveness. An example of the efficiency

of digital collaborative technologies is the production of open-source
software, which is supported by the voluntary work of thousands of
programmers around the world, who can develop very sophisticated and
competitive operating systems and specialised applications (see more on
decentralised open-source structures in Benkler, 2006).
When discussing collaborative technologies, we need to recall the concept
of positive network effects: the idea that the more people are connected
to a network, the more benefits for all in the network. We have seen this
before: the more people use the electric infrastructure, the cheaper the
service for all. In social networks, the more people are connected to the
same infrastructure, the lower the cost of communication between any two
points in the network and the lower the cost of searching for interesting
ideas and connections in the same network.
Let’s think about collaborative technologies such as SharePoint, Microsoft
Teams and IBM connections. Although these tools are proprietary, they use
the very same logic of social technologies such as LinkedIn and Facebook.
These propriety tools allow thousands of employees and associated
professionals to interact, within the boundaries of an organisation or beyond.
Each person adds their profile with the details of their past experience, their
current projects and interests, their expertise, their availability at a given
moment and their professional connections. In addition, the tools allow the
creation of discussion groups, which come together to share opinions and
information on particular topics and working groups, which are temporary
organisations for solving a punctual problem or executing a project. They
also facilitate collaboration in producing documents together.
Collaborative technologies like these examples change the way people
work together. They smooth the hierarchical structures, creating channels
for people to contact each other directly, based on their expertise, instead
of their department or localisation. They facilitate the clarification of
questions through discussion groups, the distribution of work through
working groups and the production of common documents. They also
make more visible who is contributing to different projects and enabling
innovation in the organisation.
In order to protect their intellectual property and competitive advantage,
in addition to the privacy of customers and employees, companies prefer
to use proprietary social technologies. In this way, they have more control
over what is produced and who has access to the content. However, we
can see many similarities in outcomes if one decides to combine free social
media. Indeed, people form informal groups on LinkedIn and Facebook to
discuss topics of mutual interest. They find new contacts through searching
the profile of people in these services. And they share the content they
produce individually or collectively in these services and beyond, using
SlideShare, YouTube, Twitter and blogs, to site name only a few.
Activity 13.3: How Facebook and LinkedIn support discussion groups

As discussed above, one of the main benefits of social media is to enable collaborative
interactions. In this exercise, you should research on Facebook and LinkedIn to find some
discussion groups specialising on a topic. For instance, find groups that discuss social
media marketing. Enter some of these groups and see their dynamic. Do you think these
groups may help your learning process?
151
13.2.3 The emergence of social media marketing

Organisations have increasingly been using social media for marketing. For
instance, companies use social media to increase brand awareness, have
more conversations with customers and potential customers and publish
content these audiences would appreciate. Through these interactions,
the company may know more about customers’ needs, using these ideas
to create new products and services. Companies also use social media for
advertising, for improving the reach of the right customer at the right time.
By giving better information on individual profiles and interests,
social media tools provide a better fit between the advertisement
and the customer, increasing the chances of converting the view of
an advertisement into sales. More technically, social media use the
information gathered from users to build powerful algorithms based on
profiles, behaviours, interests and beliefs. Even emotions can be added
to the algorithm to improve the fit of a particular advertisement with a
particular customer in each interaction. The better the algorithm, the
better the result for the same investment.
To understand the way companies use social media for marketing, Tuten
and Solomon (2013) propose a framework focusing on the nature of the
interaction with customers. They divide the social media channels in four
zones of activity. The first is the social community channel, in which
the company interacts with customers through sharing, socialising and
conversing. The second is the social publishing channel, in which
the company focuses on the content to be published, either produced
by the corporation or generated by customers. The third is the social
entertainment channel, in which the company engages with customers
through games, music and art. The fourth is the social commerce
channel, in which the company participates in spaces in which customers
share ideas and opinions on products; the objective here is to cultivate the
relationship with the customer, offering better service to get more sales.
Particularly concerned with the consistency of customer experience
through social media (for marketing, branding, customer service and
customer interaction), Smith et al. (2011) propose a framework to
guarantee the quality of content creation, the capacity of analysing the
data generated on social media through analytics and the provision of
clear rules of governance for guiding employees when using social media.
Based on their experiences with companies, the authors say there are six
critical success factors for companies to succeed when using social media:
context, culture, process, metrics, people and policies.
The context should take into consideration the business environment and
objectives and the regulatory environment. The culture should focus on
understanding habits and behaviours, ways of working and the subcultures
of different parts of the organisation or different segments of the audience.
The processes should be aligned with the objectives of the organisation,
having clear leadership to guide the development of this strategy. The
metrics would require the organisation to have enough technical and
human resources to measure the performance on social media channels
(such as sentiment analysis). People should be trained for using social
media as defined by the company, with clear roles and responsibilities
and best practices. Finally, the policy means the company makes clear
to employees what is expected from them when using social media,
empowering people to better use the tools within defined limits.
152
Social media functional building blocks

An alternative approach to discussing the best strategy for using social
media in customer engagement and marketing is to focus on the social
media functional building blocks (Kietzmann et al., 2011). This framework
understands that social media tools have different features and that
companies should combine these features in accordance with their strategies,
to get more benefits. This framework helps companies to better integrate
social media into the overall marketing strategy, getting organic likes,
followers and sharing and positive commentaries and ratings (earned media)
to support the efforts of advertisement and branding (paid and owned
media). In this direction, the authors suggest that social media tools have
seven key features – functional building blocks – to be analysed: identity,
conversations, sharing, presence, relationships, reputation and groups.
The identity feature measures how much the users reveal of their actual
identity. This allows the company to have a deeper understanding of
users, demanding data privacy controls. Conversations measure how
much users communicate to each other. Thus companies can measure the
speed of conversations and how conversations start and spread (diffusion
of ideas). The sharing feature measures the extent of the exchange of
content among users, allowing companies to understand the content
they share and the social graph of content production and distribution.
The presence depends on whether users know if others are available,
allowing companies to create a real-time interaction. Relationships
show how users are related to others (social network), allowing
companies to manage the structural properties of the social network (for
instance, understanding who the key influencers in a network are). The
reputation measures how much a user can know about the opinion of
others and the content others publish. Companies can then monitor the
sentiments and reactions of customers through this feature. Groups
mean users can form communities and; thus companies can understand or
influence the rules and protocols of these communities.
Mapping social media tools would help the organisation to find where
relevant communication and interaction is happening and who are the
influencers in these interactions. In the social media landscape, the
company is not controlling the conversation. At any moment, in any
node, a new piece of information may be generated and become relevant,
depending on how it spreads. The company thus needs to be prepared
for using the available analytical tools to get more insights (such as
TweetDeck, Google Alerts and Google and Facebook analytics). These
insights may also be about competitors, when the information is public.
Finally, curation of content is also important: the company needs to
know what sort of content is of the interest of the audience and when the
content is welcome.
Activity 13.4: Viral marketing

One of the key benefits of using social media for marketing is to get people
spontaneously sharing good content about a product or company. When a marketing
campaign is very successful, it can go viral through social media: lots of people share
the content with friends. For this exercise, first go to Wikipedia to learn more about viral
marketing (https://en.wikipedia.org/wiki/Viral_marketing). Second, research using Google
Search examples of viral marketing. Apply some concepts explained in this section to
better understand how these pieces of marketing have gone viral.
153
13.2.4 Risks and challenges of social media

Social media has also brought many risks and challenges to individuals,
companies and societies and we need to address these concerns with
great attention to avoid jeopardising the benefits of using these tools.
We list below some of these risks and challenges from the perspective of
organisations, with some recommendations. Considering the limitations
of this subject guide, we need to summarise the topics. Much more could
be said in terms of risks and challenges of social media for individuals and
organisations to societies.
The roots of these challenges and risks are many and most of the time
they come from a combination of factors. Some of these roots are: the
design of the tools (the technical features); the contractual agreement
between the service provider and the user (policy agreements); the way
individuals use the tool, following or not the norms defined by a social
group; the legal and regulatory framework (or the lack of those); and the
unexpected outcomes from the interaction of all factors together. Added to
this complexity, there are billions of users around the world, each one with
their own interests and views and we can imagine that new concerns are to
emerge.
Security and privacy

As with any other digital infrastructure, social media is subject to
cybercrime. Hackers may either take control of social media channels or
may use social media channels to spread harmful content, such as viruses,
Trojans and spyware. Companies are aware of these risks and in general
they have security systems in place to impede such attacks. However, there
is a particular weak link on social media: the people. Cybercriminals are
specialised in social engineering their attacks, to use psychology and what
they effectively know about people to get access to more information on
social media and from there to prepare further attacks. Companies have
special concern in protecting the privacy of customers and employees. In
addition to add more security layers, companies may also invest in training
employees to avoid socially engineered attacks.
A critical example of privacy breach has occurred when the company
Cambridge Analytica gained access to data from Facebook users without
their consent. The company Global Science Research developed an app to
identify the psychological profile of Facebook users. People would access
the app and pass their information to the company. However, when a user
answered the questions, the company had also automatic access to the
personal data of all connected friends. This information was passed later to
Cambridge Analytica, which used the personal data of 87 million Facebook
users, including their psychological profile, for the political campaigns
of politicians in the USA, such as Donald Trump, and for the Vote Leave
campaign in the United Kingdom (during the Brexit referendum) (both in
2016). The privacy of individuals was breached because Facebook allowed
the collection of data from the friends of those who accessed the apps –
without their consent. The case received public attention only because a
former Cambridge Analytica employee revealed the scheme for the data
to be taken from Facebook without the consent of users. Otherwise, those
affect probably would never know what has happened with their data.
Damage to reputation
There are many ways in which reputation can be damaged on social media.
We focus here on situations in which customers have made public their
complaints about products and services, gaining viral attention from other
154
customers. In a matter of hours, a company may have serious reputation

damages, if action is not taken quickly to sort out the problem. Sometimes
the damage lasts longer, particularly when the complaint is not answered
or when the criticism attracts many other people with similar feelings. One
interesting case is the viral criticism known as ‘United breaks guitars’. A
group of musicians made a trio of songs to express their disappointment
with United Airlines, gaining viral attention. Although the case was in
2009, the protest song is still very much alive on YouTube.
Activity 13.5: United breaks guitars

Research online about the case, understanding how social media has been used to
attack the reputation of United Airlines. Consult the Wikipedia link on the case
(https://en.wikipedia.org/wiki/United_Breaks_Guitars) and watch the video on YouTube
(https://www.youtube.com/watch?v=5YGc4zOqozo). Check how many people have
watched the video and their commentaries on YouTube. You can see how a viral
campaign can damage the reputation of a company.
Waste of time and resources

Many companies are strongly concerned that the excessive use of social
media is reducing corporate productivity. The obvious argument here is
about the distraction. People are checking their social media interfaces
many times in the day, not concentrating on the work to be done. The
more people spend time in these conversations, the higher the risk of
reducing productivity. Naturally, within limits, the conversations are
supposed to increase productivity. However, nobody has a clear rule about
how much social media use is adequate to avoid reducing productivity.
Information overload from social media is another element that could lead
to a reduction of productivity in companies.
A further question is the impact of social media tools on our cognitive
capabilities. This is a more serious concern. Some authors argue that
excessive use of social media has a lasting impact on our brains. As social
media is fragmented and multichannelled, our brains learn how to jump
from one topic to another very quickly. However, in exchange, the brains
lose the capacity to concentrate on the same topic for long periods and to
develop deep analysis. This would be the result of having our attention
span reduced because of the excessive use of social media. This is an open
discussion, but we need to keep in mind both questions: whether social
media is affecting our productivity and whether it is changing our capacity
for concentration. To find out more about this discussion, see Carr (2011).
Lack of strategy
Many companies start using social media by following others, without
carefully thinking how social media would help the organisation to get
closer to its strategic objectives. Without aligning social media with
business strategy, the organisation runs the risk of not choosing the right
tools and the right level of support in the use of tools. Companies need to
think in advance how about the coordination of efforts across departments
and units and ensuring consistency in the content which is made available
in different channels. They also should have a plan on how to collect and
analyse data generated by social media and how to make policies for using
social media clear, with rules for reward and accountability. Similarly,
professionals should have a clear strategy for individuals’ use of social
media. See more on this matter at Smith et al. (2011).
155
Lack of leadership
People may be interested in participating in social media, but they are not
sure of the benefits and risks. In corporate environments, organisations
may work to provide such leadership. First, leaders should keep social
media activity in alignment with the corporate strategy, designing proper
social media architectures (the combination of tools and features).
Second, they should use analytics to drive their actions, recommendations
and capacity for coordinating the efforts of others. Third, they should
guarantee that the content produced is relevant, that the correct people
are reached by the communication and that users have means to filter
the content in accordance with their needs. See more on social media
leadership in Deiser and Newton (2013).

This chapter brought key ideas related to how organisations are using
social media infrastructures. Mainly, the chapter focused on the link
between social media and collaboration and innovation, presenting
some conceptual frameworks to explain the phenomenon. The chapter
also focused on explaining how organisations are using social media for
marketing activities. In the last part, the chapter discussed a few risks and
challenges of social media for organisations, with some recommendations.

• explain how social media foster collaboration within organisations and
beyond
• describe the link between the idea of collaboration and the process of
innovation
• discuss key risks and challenges associated with social media tools
• apply some frameworks to understand the use of social media by
organisations.

1. Companies are using social media for fostering collaboration. Explain
how social media may support collaborative efforts in corporations.
2. Research and theories point out that social media may foster
innovation. Present arguments which support and challenge this idea.
3. Social media is an important tool for marketing, but it also bring risks
for organisations. Critically discuss the statement.
156
Chapter 14: Delivering digital products and digital services
Chapter 14: Delivering digital products

and digital services
Chapter outline
This chapter aims to further develop the discussion on the delivery
of digital products and digital services. We have already discussed
some aspects of the topic in previous chapters about the internet, the
WWW, cloud computing, mobile technologies and social media digital
infrastructures. This chapter focuses, though, on the top layers of digital
infrastructures, which are those related to providing digital products and
services.
14.1 Introduction
This chapter looks at a collection of examples of products and services
supported by digital infrastructures. Although we separate products and
services in accordance with the digital infrastructure, this classification is
somehow flawed, as any of these products and services also depend on
other digital infrastructures to operate. For instance, games are services in
social media and apps in mobile technologies, but both depend on internet
infrastructure to work and people play games through social media in
mobile devices too. As we have argued before, all digital infrastructures
discussed in this subject guide are connected to each other in layers. For
this reason, it is not possible to say that a particular product or service is
only related to one digital infrastructure.
In addition, this chapter does not aim to be either an exhaustive
description of all sorts of products and services (this would be impossible
anyway), nor a deep analysis of any of the examples. The goal here is just
to offer an introduction to key products and services, some well-known,
others still becoming more relevant. We try to keep a balance though, to
offer you a vision of the variety of products and services that are available
for individuals and organisations. We also bring examples of digital
products and services developed as a response to the Covid-19 pandemic,
which demonstrate the generative potential of digital infrastructures in
face of changes in the environment.
We want you to learn about different products and services, to understand
some current developments which may become massive infrastructures
in the future and to try to connect these products and services among
themselves. As argued in Chapter 12, the variety of apps are such that
mobile technologies has opened unlimited opportunities for businesses
to innovate just through the recombination of apps. This becomes even
more important when we see a broader horizon of digital products and
services which may be combined. Thus, we hope that this chapter shows
you a new perspective on innovation itself from a business perspective.
Being an innovator in digital business does not mean you are a developer
of software or a computer scientist. You may be a business innovator just
knowing how to combine digital products and services to get the best from
the available digital infrastructures.
157

• Introduce key digital products and digital services.
• Relate digital products and digital services to digital infrastructures.
• Highlight applications of digital products and services.
• Point out current developments of digital products and services.

• outline the key features of various digital products and services
• explain the connection between the interaction of different digital
infrastructures for providing digital products and services
• critically discuss some digital products and services
• describe how digital products and services support business
innovation.

[ISBN 9781292417752]. The student should read many parts of the
book which are related to the internet, WWW, cloud computing, mobile
technologies and social media.
Internet of Things (IoT): https://en.wikipedia.org/wiki/internet_of_things
Cloud computing: https://en.wikipedia.org/wiki/Cloud_computing
Mobile apps: https://en.wikipedia.org/wiki/Mobile_app
Meta Business Suite: https://www.facebook.com/business/help/20561413085
2988?id=765488040896522
Mashups: https://en.wikipedia.org/wiki/Mashup_(web_application_hybrid)

Cusumano, M. ‘Technology strategy and management: Cloud computing and
SaaS as new computing platforms’, Communications of the ACM 53(4)
(2010), pp.27–29.

This chapter offers brief descriptions of various digital products and
services, organised by digital infrastructures. The chapter starts with cloud
computing services: infrastructure, platform and software as services. The
chapter starts by presenting some digital products and services associated
with mobile technologies and social media, detailing later those which
are related to the internet and WWW as a whole. The chapter ends by
exploring some ideas about current developments of digital products and
services, which may become massively relevant if their potential is fully
realised in the next years.

14.2.1 Cloud computing as services
We have already introduced the definition of cloud computing services
in Chapter 11. Revise there the concepts in terms of definitions. In this
section, we present some examples and applications of cloud services.
The same providers may offer all layers of services. Organisations
and individuals may prefer, though, combining services from different
158
providers. The biggest digital companies are providers of cloud services,

including Google, Apple, Microsoft, Amazon, IBM, HP and many others.
Infrastructure as a Service (IaaS)

There are many technical details related to the provision of infrastructure
as a service. Let’s focus on the server aspect of this large infrastructure,
for a company that needs to have scalability and look for a hybrid solution
(having part of digital infrastructure on its own premises and part in the
cloud from a service provider). Think, for instance, about a company
which develops games. This company develops new games regularly, but
it is not clear when a particular game is going to become a bestseller. This
company may use a cloud hybrid infrastructure to get the flexibility it
needs in case it is necessary to scale the provision of services suddenly. In
this example, the company would have its own information technology
infrastructure and only part of it would be in the cloud provider, with a
flexible contract to be able to use more of the cloud servers when more
gamers are using the service.
The company does not need to invest beforehand in lots of servers which
may never be necessary. On the other hand, the company also does not
run the risk of missing opportunities if a game sells well. By reducing risks,
the company can be more innovative. This solution allows the company
to keep control of its main operations in-house and just adding cloud
servers when necessary. The company is prepared from day zero to scale
its capacity, as all the contracts and infrastructure which allow the uses of
the cloud resources are already in place. This sort of model has been the
choice of many companies which either do not know when they would
need more capacity, or that need more capacity only in defined periods of
the year. An e-commerce service, for instance, may need much more server
power close to special festivities, when people by presents to each other.
A cloud provider is the best solution to these seasonal picks of demand in
Platform as a Service (PaaS)

This service is specifically for application developers only, i.e. those who
use software platforms to design other software and interfaces. This
allows application developers to do their job of creating new pieces
of software and new interfaces without investing in the purchase and
maintenance of lots of hardware and software to enable them to work.
For instance, Microsoft Azure is a platform which provides programming
languages, tools and frameworks for developers to create new pieces of
software. Indeed, Microsoft Azure is a more complete service, offering
the infrastructure to host developed websites and to manage the data
necessary for the developed applications. Similarly, Google App Engine
offers a platform as a service for web applications which are hosted in its
datacentres (for small users, the service is provided free).
Start-ups may greatly benefit from this solution, as instead of using their
capital to invest in a creative platform, they can use the service in the
cloud (pay-as-you-go and) and save the money to other aspects of the
businesses. This simple change means more application developers can
start their own businesses instead of working for bigger corporations.
With this resource, small and medium businesses may enter the marked,
competing with more powerful organisations.
159
Software as a Service (SaaS)

The provision of software as a service has changed the way we manage
computing. Two decades ago, if a company needed software, it would have to
buy a licence. The licence would vary according to the number of users. Most
vendors would put a limit on the minimum number of licences to be used by
a company. This model was a barrier for SMEs, as many would not have the
resources to start using the software, either because the original investment
was too high or because the maintenance of the software (and its integration
with other pieces of software) would be too expensive in the long term.
SaaS means that companies can use the software itself in the cloud. In
order to get this market, service providers have given some benefits. One
of them is the reduction of barriers to adopt the software. Clients can buy a
small number of licences and be allowed to buy more at any time they have
more employees. This not only saves on initial investments (the company
does not need to go through difficult exercises of estimation of business
needs before starting to use a piece of software), but also saves money
across the process of using the software, as the maintenance and update of
these interfaces are also done in the cloud by the software provider. This
reduces internal maintenance costs.
Let’s pick up the example of SalesForce. The company has innovated
through offering a customer relationship management (CRM) product for
any company, even the smallest ones. Previously, the traditional offering
of CRM solutions was too expensive for SMEs. But once an SME became
big enough to afford a CRM solution, all the customer data would need
to be migrated into a new information system and many times this
migration would be very problematic (depending on the legacy systems).
With SalesForce, an SME company could start using a professional CRM
service from the beginning. The more the company grows, the more it
uses SalesForce, again in a pay-as-you-go model. SalesForce has grown a
lot since its creation and today it offers other associated cloud services,
including platform solutions (an API) for developers to create plug-ins
to the service and software interfaces (API) to manage the interaction of
companies and customers on social media.
14.2.2 Mobile technologies

In Chapter 12 we discussed some of the digital products and services
associated to mobile technologies. This section gives more details on them.
The app services

The app services are one of the most important innovations brought by
mobile technology infrastructures, not only because of their functionalities
and variety, but also because of the way people and organisations can
recombine them to get a particular outcome. The apps have added huge
value to our smartphones and tablets, in addition to making our lives
easier. Just think about buying a train ticket in an app and getting to the
train having just your electronic ticket in your mobile. Apps help us to book
restaurants and services, to get the latest news, to read books and watch
films and videos, to purchase from our preferred stores, to make payments
and to collect data on our health and activities during the day, to cite just a
few examples.
All mobile phones and tablets come with some pre-installed basic apps,
such as email client and calendar, web browser, maps and the interface to
download other apps (the apps stores). The apps today also offer the more
traditional desktop application software packages, such as Word, Excel and
PowerPoint (or similar solutions). These apps have increased productivity
160
and allowed people to work when commuting to different places. For

instance, the interval between meeting two clients may be used to answer
emails, prepare documents and collect information on the interests of the
next contact to be visited.
There are specialised apps for organising tasks and priorities, integrating
communication channels, making conference calls, organising passwords,
taking quick notations, designing ideas, integrating calendars with maps,
checking bus and train timetables, planning journeys, controlling expenses,
managing customer relationship, among many others. There are also apps
to capture how you are spending your time online, for you to analyse later
if you are wasting too much time with useless tasks and then improve the
quality of your time online. In addition, coaching apps can help you to
achieve your goals, showing how well you are sticking with new habits
and progressing towards your goals. Apps have also changed the way we
understand entertainment time, making available games, music and films
everywhere (as long as the broadband connection is good enough).
Let’s examine the example of a coffee shop chain using an app. This company
may create an app to facilitate the lives of their loyal customers. Anywhere
a person wants to find one of its coffee shops, the app uses the mobile GPS
and integrated map facilities to find and give directions on how to get to
the closest one. It may even incorporate extra information such as time for
queueing and allow online purchases so the customer can get their order
straightforwardly when arriving at the shop. The same app can integrate
information about the customer purchases and quickly communicate
repeated orders to the store. It also can give reward points to regular clients,
as the store will have an easy interface to manage the relationship with
customers. Even a small catering provider – like independent coffee shops
and family restaurants – could benefit from a similar app.
Activity 14.1: Apps from different coffee shops

Research what apps are available for different coffee shops chains: Starbucks, Dunkin’
Donuts, Costa Coffee, Pret A Manger and your local offerings. List their characteristics and
how these apps deliver benefits to customers. Discuss how you could improve the quality
of these apps.
Localisation services
Mobile technology is the perfect host for all sorts of localisation services.
The basic one is offering a map, with GPS features (finding directions). It
is a fundamental feature to help people find their way to particular places,
leaving behind paper maps. It is also a way to improve customisation of
advertisements: a mobile provider knows you are interested in buying a
new bag and informs you of a promotion by your preferred store, which by
the way is just around the corner. Location services also help tracking the
delivery of products, such as when e-commerce services estimate the time
for a delivery. They can be also used to track physical activity, to check the
route used by sales professionals, to check alternative routes when there is
traffic on main roads. One can also find friends who are nearby to spend
time together, using location services such as FourSquare.
Activity 14.2: Understanding location services

With the advance of mobile technology, location services have become more pervasive.
For this activity, link location services with broad digital infrastructures. In what way do
location services depend on a broad digital infrastructure?
161
Activity 14.3: Singapore’s transport system

Research how the government in Singapore has adopted localisation services and digital
monitoring of traffic to improve the quality of transportation systems in the country.
See how digital mobile devices are used to support the Intelligent Transport System in
Singapore.
Databases
Mobile technologies have become a point for collecting and storing data,
which are useful to support other digital applications and tools. Your
mobile devices have the technology and data to know what are your key
interests and friends, which places you go regularly, when and by which
transportations means, what time you wake up and go to sleep, how
many steps you take in a day and even your heart rate. In some cases you
need to set up the mobile device for collecting the data, downloading the
specialised apps and allowing the data to be stored in a place for further
analysis, but not always. Your basic location, for example, is known by
default through the phone network.
There is currently a particular interest in collecting data with the help of
wearables, such as digital watches, wristbands and glasses. These devices
can help collect data from your movements and activities, for instance,
measuring the number of steps you have walked, the time you have run,
the total calories you have burnt in the day, how many hours you have
slept and so on. In addition, mobile phones have apps for you to add other
data, feeding a larger database which can be used by your doctors. Indeed,
these databases are powerful tools for us to know more about ourselves
and make informed decisions about changes which may be useful in
our lives. Based on the collected data, fitness apps may, for instance,
recommend new activities and exercises, gradually helping you to become
fitter.
Activity 14.4: Omnitouch – Wearable computer

How about being able to project your computer interface at any place and time and
manipulate the interface in any surface as if it was a touchscreen? Research the state
of art of this Omnitouch technology. How business can use this technology to improve
productivity?
14.2.3 Social media

We have seen earlier that social media are opening space for game
developers to sell their products. This section gives a few examples of how
social media has become the channel for the distribution of games. We
also discuss how social media has become a provider of data analytics for
marketing.
Games
Digital games are a global success. They are everywhere. We have already
seen that many mobile apps provide games. Digital games may be accessed
from different channels, including social media. This happens because
social media facilitate invitation from friends to play the same games
together. If you like a game, the chances are that you want to share this
with your friends. Some games even use this human trace to leverage your
position in the game: the more players you bring to the game, the more
points and rewards you have. There is thus an immediate synergy between
social media and games.
162
An example of a game developer is Zynga, an American company which

develops social games to be played on mobile devices, on the internet
and on social media, such as Facebook and Tencent QQ. For instance, in
2009 Zynga created FarmVille, a game played on Facebook. The company
has grown strongly based on its presence on social media, although the
partnership with Facebook has now ended. Zynga and other similar game
companies make money through direct sales of their games, or through
advertising.
Analytics
Knowing your customer is a critical factor in business success. Companies
have always relied on surveys and demographics to know about customers
and potential customers, to better target advertisements and to better
develop products and services. Social media offers a short cut to the
process of better knowing your customer. Let’s think about Facebook, which
is the most relevant social media service. Facebook has crucial demographic
data of more than 2.9 billion people. But more important than this, it
knows with a high level of precision what people like, what are their more
recent interests, what that are talking about, who are their friends and
even how they feel today. Facebook’s powerful algorithm collects individual
data and adjusts the offering of ads in accordance with individual profiles.
Search for a product online that you have never been interested in before.
Now open Facebook and see which ads you are shown.
In addition to offering services to better target potential customers,
Facebook also offers advertisers the tools for receiving the data analytics
of their advertisements. With this tool, companies can check whether a
campaign is doing well and decide in which campaign they are going
to invest more. Data analytics allow companies to see how people are
reacting to the advertisement, with emoticons, sharing or commentaries,
on the top of the number of viewers. Based on the feedback, companies
can adjust their target audience by country, region, age and even by
interests (through key words). Other social media offer similar services.
14.2.4 internet
The internet is the core digital infrastructure that enables almost all
digital products and services. In this section we focus on a few classes of
digital products and services that have been significantly reshaped by the
internet.
Books, music, videos, films, newspapers, magazines

Once upon a time, all these products were only physical objects. Today one
may have a physical copy of these products, but increasingly people are
migrating to the digital copies. Convenience, price and social norms are
behind the migration to digital products. Ease of use is also one reason,
because of the convergence of technology: these products can be made
available all in the same device, which does not occupy much space.
Integrated with cloud services, the digital products also are available
through any digital interface and can be recovered at any time.
Search
Searching is one of the most important services of the internet – it lets
us ‘find’ to ‘connect’ and ‘find out’. Much of this is powered the WWW
infrastructure and search engines, but also by social media posts. Antivirus
services complement the search services, for example checking whether
websites are safe.
163
The most important search service in Europe and America is provided by

Google, but there are others such as Yahoo! and Bing. Other countries
around the world have their own dominant search engines. The power of
Google Search comes from the quality of its algorithm, which considers
the content and relevance of the pages and the interests of the searcher.
The more one uses Google Search, the more the tool knows about one’s
interests and the better the service can match the search with correct links.
The relevance (ranking) of the pages depends mainly on how independent
sources connect to particular content. Thus, the search depends on the
WWW to verify the way different content pages are connected to each
other and to deliver the links to the searcher.
Remote desktop control

In the old times, if one had a problem with a computer, the solution would
be to have a technician coming locally to investigate the case, or to get the
computer to the technical service. Even now many computer problems can
only be sorted in this way, especially when the defect is in the hardware.
However, for software problems, increasingly companies provide remote
support, through the remote desktop control applications. These interfaces
allow a technical professional from anywhere in the world to access your
computer online and try to fix the problem remotely. This service is used by
corporations to provide support either to the computers in their facilities
or for the users’ own devices, wherever they are. This kind of service has
increased productivity and it is also a very convenient and cheaper solution
but, it is also a service that can be abused in many ways, allowing criminals
to ‘hijack’ or kidnap your computer or your data – perhaps demanding a
ransom or perhaps using your device as part of a malicious botnet.
Mashups
A mashup is any application which uses content from diverse digital
objects, available on the internet, to create a new product or service which
aggregates information from different sources to reach a particular the
mashup interface, the audience see one final object, but behind the scenes,
there are layers which are combined to provide the final result. A typical
example is a hotel chain which combines the address and images of the
hotel to Google Maps, offering customers better information of its location.
The mashups depend on the capacity of integrating different digital
objects. It is easier to do this integration when the developer has the APIs
(Application Programming Interfaces) of the platforms used (see more
details on platforms in Chapter 18). Mashups may be used by combining
free services and information – although ‘free’ is usually contested idea
online – or they may make payment for the use of third-party content.
14.2.5 Further developments
Internet of Things
The internet of Things (IoT) is one of current developments of mobile
technology and digital infrastructures. The IoT means that digital
connectivity is associated with any physical device – from machines and
domestic devices to buildings or clothes – and that these digital elements
can communicate with the digital elements of other physical devices. Any
object can be connected to others, creating a network of communication
among objects which can send and receive data automatically. The digital
elements in each device are able to collect and distribute relevant data and
to respond to the information that they receive through the network. This
can allow remote control of individual objects, but also the pooling of data
164
from multiple objects. These digital devices associated to things are able to
collect data to be used in the management of processes and problems. For
instance, adding biochips to farm animals may allow a farmer to manage
each animal’s specific diet and the whole farm better, gaining a level
of information which could not be captured otherwise. Digital devices
associated with machines, such as cars and airplanes, may measure
the performance of different parts of the technology, collecting enough
data to provide proper maintenance before they break. You can see the
immediate advantage of being able to collect earlier warning indicators
from machines, reducing the negative impact of having failures in critical
moments. The data collected by IoT contributes to the formation of large
datasets about the behaviour of materials and equipment. These big
datasets are used then to investigate patterns of behaviour by AI solutions
(e.g. machine learning).
The IoT is already very close to many of us, through mobile wearables.
When you have a digital watch tracking your activities, this means you
have a digital interface to transform your activity into data. The data
is then transferred to an online database, for you (or your healthcare
provider) to analyse it and make decisions on your health and physical
activities. You become a Thing in this big Internet of Things when you add
a wearable to measure your performance. On a larger scale, engineers are
applying the IoT to entire cities: the smart city movement (see more in
Chapter 19).
Activity 14.5: Digitalisation and the IoT

Research how the IoT is changing the management of the servicing of airplane engines.
What are the ‘consequential changes’ that you would expect from this specific act of
digitalisation?
Activity 14.6: IoT for Smart Homes

Look for examples of how IoT devices can create a better home environment – e.g. in
smart homes.
Do you want to live in a smart home where your every act is monitored and responded to?
AI as a Service
Gradually, AI technologies are advancing. We are entering the era of
cognitive computing. These are technologies which somehow can ‘think’
by themselves, meaning they are able to mimic human cognitive functions,
including learning processes (machine learning). There is a long way to
go for such technologies to be able to do exactly what humans do and
some do not believe machines are ever going to have the same general
intelligence that humans have. On the other hand, machines are already
doing amazing things that humans alone cannot do. Interesting examples
are IBM Watson and Google DeepMind. These are powerful pieces of
software able to processes huge amounts of information from structured
and unstructured databases. Even more interesting: they can learn through
asking questions and comparing the results with other situations. They can
also generate new ideas that nobody has programmed in their system.
IBM has prepared Watson to be an AI platform. As such, Watson can
learn anything. Give Watson enough information and time and the
technology will learn anything (well… almost). The early application
for Watson has been in medicine and research related to medicine. After
being trained, Watson can now support many medical activities, such as
165
making diagnoses and finding new possible lines of research. This is an

amazing outcome. In addition, Watson is learning every day about the very
large topic of medicine but also about the processes of learning, through
interacting more with humans.
Anyone with the resources can have access to Watson, for instance, to
improve the services in a hospital or to enable better research in the area
of medicine. In other words, IBM is offering its AI machine as a service
through the internet and mobile technologies. Today the service is limited
to a few topics that Watson already has some understanding. But the list
of topics is increasing. In addition, IBM has concluded that alone they
are not able to innovate all the things that are necessary for Watson to
become a bigger AI platform. For this reason, the company has created an
API to allow software developers to use the information, knowledge and
reasoning available through Watson in a variety of apps. This might in
time allow the spread of AI as a service in any computing interface, from
desktops to mobile devices.
The potential of using AI in the cloud (AI as a Service – AIaaS) is
significant for companies and governments and has potential to add
significant value to society as a whole. For instance, Company A can
develop an automated transcription service with high quality, using AI.
Then other companies can contract Company A for providing transcriptions
for recorded videos. The process of creating transcripts might be done
100% in the cloud, giving other companies the benefits of access to the
transcription. The same Company A, using its capabilities in the area
of voice recognition, might be able to sell services in the cloud for any
company which want to automate calls to a customer relationship channel.
The buyers of the service, in this example, do not need to know anything
about AI. They just know that when the customer calls, the AI software is
going to recognise the words and guide the customer to the right assistant
(which may be a person or an automate response). We can foresee a
future in which powerful personal assistants such as Alexa, Siri or Google
Assistant will be available in the cloud for any company to contract their
services. For instance, a university could contract these personal assistants
to help students to find the content they are looking for in a library or
on the internet or in a collection of videos. We are not yet there, but the
AI algorithms are becoming better to identify questions and answers and
companies specialised in AI are also better in designing business models
which allow them to sell their intelligent services as cloud services.
Activity 14.7: IBM Watson ecosystem

Research online the companies which are using IBM Watson to provide services through
apps and other digital services. Think how IBM benefits from opening the Watson APIs
for developers to use Watson cognitive systems in their apps. Discuss the benefits for the
developers of apps and for users too
14.2.6 Covid-19 pandemic and digital services

During the global Covid-19 pandemic, millions – perhaps billions – of
people worked from home. We also know that the purchasing habits of
people changed, resulting in more traffic online and more reliance on
digital infrastructures of all kinds. It is also the case that governments
around the world turned to the internet, WWW and mobile phones to
build new information resources to help combat the pandemic. These have
often been undertaken as rapid development projects that piggy-back on a
local installed base, for example on existing databases, healthcare facilities
and, of course, networks and smart mobile devices in peoples’ pockets.
166
Also, in education, all kinds of student from children in kindergarten to

postgraduates moved from face-to-face classroom learning to various
forms of online learning.
This experience presents a very rich real-life experiment into what digital
infrastructures can do, what they can do well and what they struggle to
do. Many aspects of this huge shared global experience of ‘going digital’
are, at the time of writing (early 2022), hard to give a final judgement
about.
For example, has the largescale move to teaching online severely damaged
the educational attainment of a generation of students or does online
teaching work well? Do young people adapt better than we thought to
online learning and do their teachers as well? Have business organisations
whose employees moved to working from home (WFH – a Covid-era
acronym) seen a drastic drop in productivity or creativity, or has it meant
that people work harder and in a more focused way?
We cannot definitively answer these questions, but we can enter the
debate based on the kinds of theoretical knowledge that this course
introduces. For example, when thinking about these issues we might find
a use for concepts such as network externality, governance, net neutrality,
transaction cost theory or generativity. These kinds of ideas can give us a
language with which to analyse and discuss possible outcomes.
As one example, cybercriminals have used the opportunity of home
working, home learning and online entertainment to spread viruses,
Trojans and spyware and other sorts of attacks. Using the theory of
network externalities, we might explain the nature of the problem and
suggest ideas on how we could have reduced the problems. Cybercriminals
spotted the unique opportunity they had during the pandemics, with
people feeling stressed and isolated at home, with less support or
resources, thus being more vulnerable to attacks.
The mass migration to video-conferencing tools (such as Zoom, Microsoft
Teams, Google, Facebook, WhatsApp and WeChat) has often slowed
down the internet (loaded it with traffic) particularly in the ‘last mile’ of
connectivity. This may be made worse when, at the same time, people
at home have been increasing their consumption of gaming services and
video streaming, putting even more pressure on available bandwidth.
From a governance perspective this might suggest a need to regulate
traffic differently, to re-think net neutrality. Would it be sensible or
practical to favour work-related communication (Zoom) instead of
entertainment (Netflix)?
People have changed their purchasing habits too, turning to e-commerce
platforms and shifting their purchasing towards online forms of
consumption. There is no doubt that some e-commerce and entertainment
platforms (Grab, Amazon, Netflix) have been very successful during
lockdowns. But the question remains, are these new users of e-commerce
or streaming going to remain online or are these users going to go back
to normal (e.g. high street commerce and visits to the cinema) after the
end of Covid-19 restrictions? Possible answers to this question could draw
on the concepts and rationale of transaction cost theory, generativity and
the economics of digital goods as well as institutional theories about how
norms and attitudes are formed and changed.
167
Activity 14.8: Covid-19 developments

As an exercise, think through the following Covid-19-related scenarios and suggest what
kinds of analytical concepts drawn from this course you might want to use in discussing
the core ideas. These exercises help you apply- the thory to practice in a series of events
which have marked a generation.
Governments have developed mobile apps to help people and businesses to better
manage the Covid-19 pandemic. Examples would include, online ‘test and trace’ apps to
log your location and if you have been near a known Covid-19 carrier, online booking for
vaccination appointments and digital Covid-19 vaccine certificates. There has been a lot
of more general information resources and travel-related portals and registration systems
developed too. Make a list of all such systems that you know about in your country and
the basic architecture of these systems including their ‘installed base’. Identify the digital
infrastructures that they have used and are built on. How joined up are they? How well
did they seem to work? What conceptual tools and performance data would you see as
appropriate to analyse these systems as a whole?
In the face of Covid-19, many educational institutions have migrated their operations
to virtual learning environments (VLE). They have adopted platforms such as Microsoft
Teams or Zoom to deliver lectures and classes, which are then combined or integrated
with Moodle, Blackboard, Google Classroom and other VLE environments. Explain how
an educational institution of your choice has used VLEs and video services to keep their
operations going (e.g. educating students, arranging exams, building community). Analyse
their strategy and assess its level of success using the theory of the layered model and the
concept of generativity (Zittrain 2009).
Video streaming services, such as Netflix, e-commerce platforms such as Amazon and
all kinds of supermarket and food delivery services, have faced an increase in demand.
In the case of consumer goods and food this includes demand for physical delivery. This
has required these kinds of company to innovate – to bring new ideas and a new scale
to their operations. Some organisations have been able to do this, some have struggled.
Using one or two examples familiar to you in your local context, explain theoretically how
the companies may have benefited (or not) from using a cloud-base infrastructure and
services, mobile technologies and social media.

This chapter described a variety of digital products and services, in
accordance with the key digital infrastructures which are the focus of this
subject guide. The chapter emphasised, though, that these digital products
and services are dependent on more than one digital infrastructure, thus
the classification into categories is flawed, despite being useful. The
chapter also highlighted the connection between digital products and
digital services and business innovation, aiming to motivate you to see
opportunities of becoming an innovator through a better understanding
of the available digital resources. Finally, the chapter presented examples
of products and services created on the top of digital infrastructures as a
response to the Covid-19 pandemic.

• outline the key features of various digital products and services
• explain the connection between the interaction of different digital
infrastructures for providing digital products and services
168
• critically discuss some digital products and services

• describe how digital products and services support business
innovation.

1. What is the difference between a ‘service’ rather than a ‘product’? Is
this distinction useful? Illustrate your answer using examples from the
cloud, from streaming media and from the IoT.
2. Explain why social media digital platforms have added gaming
functionality and APIs to their service? Suggest one other extra
functionality that they might try to offer and justify why they should.
3. Critically discuss the ways in which digital products and services
can support business innovation (Perhaps refer back to ideas of
generativity or Tidd and Bessant’s four modes of innovation introduced
in Chapter 2 to structure your answer, or perhaps the resource based
view (RBV) of the firm in Chapter 9.)
169
Notes
170
Chapter 15: Governance and regulation of digital infrastructures
Chapter 15: Governance and regulation

of digital infrastructures
Chapter outline
This chapter aims to develop understanding of the governance and
regulation of digital infrastructures. The objectives of the chapter are to
explore some international regulations which affect digital infrastructures,
with particular focus on the internet and its governance and regulatory
bodies. The chapter explores alternative forms of governance and
points out some challenges in face of the difficulties to regulate digital
infrastructures.
15.1 Introduction
The layered model has been used in this subject guide to explain how
digital infrastructures evolve from fragmented pieces to actual products
and services which support contemporary economies and businesses.
We can use this idea of layers also to better understand governance and
regulatory aspects of digital infrastructures taking two perspectives.
The first is that governance and regulations are themselves layers of
digital infrastructures, as they affect the way the other layers are going
to interoperate and affect the way people and organisations are going to
use the other layers. The second is that governance structures of digital
infrastructures evolve through time as new uses and new users occur.
In the same way digital infrastructures evolve through the addition and
change of its associated elements within layers, the rules for governance
and regulations also evolve with shifting technologies and altered practices
of use. Nobody could have thought from the beginning how to guarantee
the governance of the internet because the network was always going to
evolved in ways nobody could foresee. And today, nobody really knows
how current digital infrastructures are going to evolve from now and the
governance and regulation that will be necessary.
There is a fundamental reason why we should care about the governance
and regulation of digital infrastructures. In a nutshell, these infrastructures
have become a fundamental part of the tissue of our economies and lives.
The more businesses and society are dependent on digital infrastructures,
the more we need to be sure that these digital infrastructures do not go
out of control. The fact is that digital infrastructures have become critical
for the wellbeing of societies and for producing wealth. We cannot run
the risk of having such resources being misused or abused. For instance, if
governance rules are not in place, a country may use the internet to attack
the digital infrastructure of another country, in a cyberwar. Indeed, we are
already in a disguised cyberwar: daily, companies and governments suffer
systematic attacks from criminals, some sponsored by governments, trying
to steal research, commercial data and security and defence information.
What would happen if a country could not use internet resources for
one week? And one month? What happens if a country has its mobile
communication system stopped by another government?
This chapter develops topics which can help our understanding of why it
is necessary to have governance and regulation of digital infrastructures,
to protect critical services in our economies and societies. However, it
is also necessary to be critical of how governance and regulations are
171
going to be used, in order not to destroy the benefits of having open

digital infrastructures which can be used by all. Any sort of governance
means that some international and national bodies will have more power
than others. Thus, this discussion is also about who are the international
bodies and systems to be trusted to protect the interests of people. We
do not have an answer to these questions, but we want to share with you
concerns and interesting questions on the matter.

• Conceptualise governance and regulations of digital infrastructures.
• Link the hourglass architecture to the analysis of governance and
regulation of digital infrastructures.
• Give examples of governance and regulation aspects to digital
infrastructures.
• Introduce key governance institutions of the internet.

• explain the relevance of governance and regulations in the way digital
infrastructures operate
• discuss the application of the hourglass architecture to the governance
layers of digital infrastructures
• outline the key arguments in some polemic discussions on governance
and regulations of digital infrastructures
• describe the relationship between the governance and regulation
aspects of the different layers (hardware, software and content) of

Internet governance: https://en.wikipedia.org/wiki/internet_governance
Internet Governance Forum: https://en.wikipedia.org/wiki/internet_
Governance_Forum
Internet Engineering Task Force: https://en.wikipedia.org/wiki/internet_
Engineering_Task_Force
Internet Society: https://en.wikipedia.org/wiki/internet_Society
W3C Standards: https://www.w3.org/standards/

The Internet Society: http://www.internetsociety.org/
Open Mobile Alliance (for discussion on mobile standards and regulations):
http://openmobilealliance.org/

Benkler, Y. ‘From consumers to users: shifting the deeper structures of
regulation towards sustainable commons and user access’, Federal
Communications Law Journal 52 2000, pp.561–79.
Constantinides, P. and M. Barrett ‘Information infrastructure development and
governance as collective action’, Information Systems Research 26 2015,
pp.40–56. http://dl.acm.org/citation.cfm?id=2895539
DeNardis, L. ‘The emerging field of internet governance’ Yale Information
Society Project Working Paper Series (2010). Online resource:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1678343
Lessig, L. Code version 2.0. (New York: Basic Books, 2006)
[ISBN 9780465039142]. http://codev2.cc/download+remix/Lessig-
Codev2.pdf
172

This chapter extends the discussion on digital infrastructures to the
level of governance and regulations. The key argument of the chapter is
that governance and regulations are layers of the digital infrastructures,
affecting the form these technologies are used. The chapter brings
examples of governance and regulations, focusing on the internet
digital infrastructure. It points out, though, other aspects that affect the
governance and regulation of cloud computing, mobile technology and
social media. The objective of the chapter is not to cover in depth the
regulations related to each of these digital infrastructures, but to offer
a conceptual perspective on the relevance and nuances of the debate.
The activities of the chapter are designed to guide the student in further
research of the topics.

15.2.1 Conceptualising governance and regulation
Governance and regulations are terms related to the establishment of
principles, norms, rules and procedures which allow organisations and
institutions to better manage a particular process or a particular asset. The
aim of having governance and regulations is to get better efficiency and
reliance in relation to expected outcomes and to reduce the risks involved
in the management of processes and assets. Corporations, for instance,
have governance mechanisms and processes in order to control and direct
the way things are done and ensure that decision making is legitimate
in its environment. For instance, the governance rules should define the
corporate policies and inform who is involved in decision making and how
the organisation monitors and controls the actions and practices of its
employees, directors and service providers. Corporate policies should be
based on broad laws and regulations, to be sure the organisation is doing as
good a job as possible to protect the interests of its stakeholders and society.
There is, however, a big difference between defining the governance
of a corporation – which is already a complex task – and defining the
governance of a digital infrastructure which operates around the globe.
Trying to control the development of digital infrastructures centrally is
almost impossible; businesses, governments and individuals will find
their way to not follow imposed central rules, thus bringing unexpected
outcomes for the whole infrastructure. Thus, a bottom-up approach may
be more effective in bringing social actors together to build a public good
which is of the common interest of all (see more in Constantinides and
Barrett, 2015).
Think again about the hourglass architecture. We have argued in this
subject guide that digital infrastructures are not designed from scratch.
Rather, they evolve through time through the interaction of many layers.
In such a model, any digital infrastructure is not coordinated centrally by
any institution. Each part of the big assembly which results in a digital
infrastructure has its own history, with legacy systems (installed base)
not only in terms of hardware and software, but also in terms of the
institutional and regulatory bodies which support different elements of
these layers. This is the task behind discussing governance and regulation
structures for digital infrastructures which have become critical to our
societies.
173
15.2.2 Internet governance and regulation

Let’s focus on the governance of the internet as a critical digital
infrastructure. The internet is a global distributed network which is not
owned by any person, or organisation, or government. It is a collection of
individual networks which voluntarily connect themselves to the whole
through communication standards. However, for each network to be able
to join the internet, and the internet to be operational, it is necessary to
create norms of governance and regulations, which are adhered by each
of the networks in a voluntary basis, to have in exchange the benefit of
being connected to the whole. For communication to flow from any point
of the internet to the other, the governance rules and regulations should
be respected by all parts involved in this communication, at least in the
fundamental levels of communication standards and protocols.
Imagine a government which decides that a country is not following the
established governance and regulations of the internet in terms of internet
protocols (software layer). The first outcome would be that the network
in this country will not be able to connect to the other networks. However,
even if the hypothetical country would be able to connect to the internet,
the other nodes of the network would not be interested in connecting
back, depending on the sort of damage that this particular network would
bring to the whole. In other words, the other nodes could decide to isolate
a particular country if the governance rules were not respected. Thus,
although the adhesion to governance rules is voluntary, in the case of the
internet, the decision not to follow the rules may have a higher cost of not
been allowed on the internet.
To work as an integrated system – although it is indeed a collection of
parts which are interoperable – a lot of institutionalised structures need
to give support to a particular form of governing a digital infrastructure.
In the case of the internet, civil society bodies, corporations, government,
and academic and research institutions come together to cooperate and
agree on a particular governance model. For instance, if you want an
internet address for a new company, you need to have a domain (an online
address). You cannot go online, for instance and publish content in the
name of www.coca-cola.co.uk or claim the property of this domain. First,
because The Coca-Cola Company is already using this domain. Second,
because, legally speaking, you are not the owner of this trademark.
If the internet is free and people can do much of what they want, how is
this sort of abuse avoided and controlled? It is possible only because there
is a global institution called ICANN (Internet Corporation for Assigned
Names and Numbers), which registers domains and controls who has the
legitimate right for a domain. The domain is a unique identifier of the
content of an organisation or individual. If many people could use the
same domain, the internet would be useless, as nobody would be able to
find content securely and search engines would not work properly.
ICANN is a global organisation directed by professionals who represent
corporations, government and civil society. However, ICANN Affirmation
of Commitment is defined by the US Department of Commerce and the
US governmental body. This situation has its roots in the way the internet
has evolved: the internet was in the USA and since the beginning has been
regulated by the US government. The USA was the first country to create
the domain rules. Once other networks decided to join the US model,
they also accepted the rule of ICANN, which, in addition to the domains,
also regulates internet protocol addresses, application port numbers
in the transport protocols and other technical matters. There is a path
174
dependency in this process, as the original governance bodies are still

working in a much more extended network.
From this situation, there is plenty of debate as to whether ICANN should
be a truly independent institution controlled by all countries. However,
there are many issues within this discussion. For some, the USA is doing a
very good job of providing the rules of governance for the internet, with
a very democratic body of institutions influencing ICANN. Why bother
to change? Others would say this arrangement gives too much power to
one country and that today we should have a collaborative attitude, but
tomorrow we may decide to change the rules of the game. On the other
hand, the US government may not have any incentive to change the rules.
The global internet depends on ICANN, which empowers the position of
the USA in the whole game. A government may decide to isolate itself
form the internet to define its own rules on its own network (see the North
Korean model). However, collectively, most people and organisations want
to be in the place others are (positive externalities), so they play the game
following ICANN rules.
Activity 15.1: Understanding ICANN and the domain system

Research the rules of registering a website domain in your country. See how your national
system is connected to the ICANN governance structures. Check also the costs of
maintenance of a domain.
The technical aspects of the internet and the standardisation of the

internet core protocols (such as the IPv4 and IPv6) are defined by another
institution, the internet Engineering Task Force (IETF). This is a non-
profit organisation which gathers professionals from all around the world
to work on particular topics and define by consensus (or the closest to
a consensus possible) what are to be the new internet standards. The
objective of the IETF is to have technical professionals who can contribute
on the definition of internet standards, including activities such as testing
the proposed technical developments. This organisation works as an
open community of voluntary professionals, who are, however, mainly
sponsored by corporations and governments. Although the IETF was also
created by the US government, since 1993 it has been managed by the
Internet Society, which is an international non-profit organisation focused
on providing leadership on internet related topics, such as standards,
education, access and policies.
Also, the World Wide Web Consortium (W3C) defines the standards of
an Open Web Platform which supports the development of applications
for the internet. The idea is to have a collection of standards for different
types of data and application which allow developers to create rich
interactive experiences for the final users in any sort of device – from
computers and mobile devices to printers, television and cars. In addition,
the W3C aims to provide an environment in which the access to the
internet resources is universal, meaning that anyone who wish to publish
on the internet should have the means, independently of the sort of device
and software they use and the sort of internet access they have. These
W3C standards define, for instance, the use of HTML5 language, of the
Semantic Web stack, XML (Extensible Markup Language), Ajax and many
APIs (application programming interfaces). The W3C works through
consensus building to make technical recommendations on software
development for the internet, providing governance and regulations
through the definition of standards.
175
In addition, there is the Internet Governance Forum (IGF), created in 2005

by the United Nations, with the objective of having a multi-stakeholder
body – again organising the collaboration of governments, corporations,
civil society, academics and researchers – which focuses on defining
internet public policies. The IGF can propose topics of interest and make
recommendations, but it cannot impose a decision on any party. One
of its concerns is to work in capacity building, for allowing developing
countries to be more participative in the internet governance. The
organisation discusses topics such as the openness (including human rights
and freedom of expression), security (hacking and cybercrime), privacy,
diversity and access of the internet. It also discusses further issues, such as
the role of the internet in fostering development, how to improve digital
trust and how to understand the relevance of the internet economy. This
body has successive mandates from the United Nations, which decides
whether or not to keep the structure working.
Activity 15.2: The Internet Governance Forum

The United Nations sponsors the internet Governance Forum for debating some crucial
aspects related to the impact of digital infrastructures on society. However, the forum may
only make recommendations. Find out online are the key recommendations provided by
the forum.
15.2.3 Governance layers

From the discussion above, it becomes clear that the governance of the
internet is not only about its technical infrastructure and the regulations
needed to make the technical aspects to work well. Yochai Benkler
(2000) argues that internet governance has three layers: the physical
infrastructure layer (hardware), the code and logical layer (the software
and the architecture, which control the internet as a digital infrastructure)
and the content layer (which refers to the information which is available
on the internet). The content layer is the most obviously controversial one,
as there is a wide range of interpretations about how much and what kind
of information should be freely accessible on the internet and other digital
infrastructures. But governance is by no means limited to the content
layer, see for example the discussion on net neutrality in the next chapter,
which expands this discussion.
Some authors have argued that internet governance needs to go beyond its
technical aspects, entering also the levels of the outcomes one may have
for using (or not using) the network. From this perspective, it is necessary
to consider the legal, economic and sociocultural aspects of the network,
exactly because of its relevance as a critical infrastructure of economies
and societies (see more on DeNardis, 2010). Some would like internet
governance to include discussion of the content of the network, which
has become a place also for cybercrimes and copyright infringements.
However, any attempt to control the content of the internet may lead
to censorship and surveillance, which are against the ideals of internet
activists, who want the network to be a place of freedom.
In terms of content, there is plenty of disagreements on how to regulate
the internet. Lawrence Lessig, for instance, has advocated the reduction
of copyright protection for content that can be distributed through the
internet. As discussed earlier in this subject guide, digital goods can be
copied at a marginal cost close to zero. Although the cost of producing
the first copy of a digital good may be high, for instance, producing a
film, once the film is done and digitalised, it is very cheap to produce
176
millions of copies of the same product. Copyrights thus create an

obstacle for the illegal copying and distribution of protected content. By
restricting reproduction, copyrights protect the right of the producer to
sell their product. But how much protection is fair from the perspective of
consumers and from the perspective of the benefit for society as a whole?
When people remix copyright content, for instance, to create a different
thing, should this be protected by copyright? Should academic research,
which is sponsored by society, be offered freely on the internet or should
it be protected by copyrights? We can see that this discussion is not
straightforward: there are many stakeholders with divergent interests that
need to be included in the debate.
Activity 15.3: Information wants to be free

There is a conflict of interest between those who produce digital content and those who
consume it. There is a debate as to whether internet content should be protected by
straightforward copyright laws, or whether we should be more flexible by establishing
tolerances for fair use of content. Research this debate online and define your own
position. You can start looking for Lawrence Lessig’s content online.
Access to content can also be filtered by governments, which may disagree

with the overall governance of the internet on this matter. For instance,
the Chinese government prohibits some sorts of content and some services,
such as Facebook and Twitter. European governments have strong rules
regarding hate speech. The Iranian government prohibits the search of
key words which can allow a citizen to reach the content of opposition
movements. The Saudi Arabian government controls content it considers
improper for the country’s culture and religious beliefs. Governments may
use ISPs to filter or control what people can access online.
At the same time, many governments around the world have made efforts
to stop illegal content being accessed by individuals in their countries,
often relying on the good will of search tools and service providers to filter
the content or to inform authorities on suspicious use of the internet. For
instance, many citizens would expect governments to be able to investigate
the internet communication and traffic to restrain and avoid cybercrimes.
They also expect governments to be able to investigate the internet to
find criminals and terrorists who use the network for communication and
organisation. However, in order to deliver these expectations, governments
need to increase the surveillance of all, with reduction of privacy. There
are trade-offs whatever regulation a society chooses.
Sometimes the regulation of the internet comes from the content or service
provider. An interesting example is Facebook. The social network defines
in its policy what users can publish on the network. When the policy is
not respected, Facebook deletes the content independently of the will of
the user. Even worse, Facebook usually punishes the user suspending the
account for a period, or just allowing the user to see the content but not
to publish or to comment on others’ walls. Is this acceptable? Some would
say Facebook is a proprietary service, thus the company can decide what
content is acceptable. If a user is not happy, just change to another social
network which has different rules.
However, others would say that Facebook has become too dominant to be
allowed to make decisions alone, without further regulation. For instance,
Facebook has entered polemic disputes with users for deleting photos
of paintings with naked models. The painting can be in a museum and
anyone of any age can see it in the museum as art, but no one can publish
177
the same painting on Facebook. Should society have a say on this matter?
In trying to be a universal service for anyone in the world, Facebook has
created policies which are not consistent with the freedom many have
in democratic countries. Also, Facebook policy is imposed top-down,
not taking into consideration that national laws may protect freedom
of expression and speech of citizens who want to publish, for example,
artistic content of naked people.
Activity 15.4: Facebook censorship

We have described above how service providers such as Facebook can create their own
rules, overriding the rights people take for granted in terms of freedom of expression.
Research Facebook’s policy. Do you agree with it? Have your read it before? Do you
know that at any time Facebook can change its policy? How do you think governments
could influence or regulate the policies of huge global service providers in digital
infrastructures?
15.2.4 Self-regulation and industry regulation

Governance rules and regulations are not neutral. There are always
interests behind the technical choices, which influence the way we
behave online and how the content we publish can be, to a greater or
lesser extent, controlled by others – companies and governments. Lessig
(2006) puts forward the argument that code is law, meaning that the
code which supports the internet is also an instrument of controlling
society. We can see this idea of code as law in a broad way: even when
not used for surveillance and control of society, a technical device is
going to influence behaviour. We cannot avoid this, just as we cannot
avoid having our behaviour influenced by other pieces of technology.
The point is to guarantee that the governance of the internet is done
in the most democratic and transparent way. At least in the code layer,
it has been more viable to have self-regulation of the internet, through
the participation of a diverse range of stakeholders in the definition of
standards (through ICANN and the IETF, for instance). In the hardware
layer, countries have more to say, although this is not as controversial
as the content layer, which is the most vulnerable layer in terms of
governance and regulation.
In the hardware layer, industries have been instrumental in defining rules
of governance and regulations. At the end of the day, the economic actors
in an industry want to get business done. Imagine a world, for instance,
in which each mobile operator used a different standard to communicate,
in such a way that a customer of Network A could only talk to other
customers of the same network. Certainly, this fragmented network would
not be as useful to the customer as the current model. This is possible
because economic actors in the mobile industry cooperate to create
common regulations which are shared by all. They leave competition to
the sphere of product and service creation and provision, agreeing though
that for the good of all, it is better to have a common standard at the
communication level. Some would see this model as the best way forward
in terms of reaching proper regulations for the hardware layers of the
digital infrastructures: the industries should define the regulations of each
aspect of the digital infrastructures, avoiding the influence of governments.
This consensual model among industry parties may work well for some
aspects, but they may also end in irreconcilable disagreements. For
instance, in some countries, mobile operators had a ferocious dispute with
customers for years, regarding having to change mobile number when
178
switching service providers. This sort of disagreement between industry

parties is not in the interest of customers and governments have been
helpful in forcing mobile operators to change their behaviour.
The consensual model also works for some aspects of the software layer.
For instance, agreement over using HTML language and the standards for
adopting WWW infrastructure. However, code is also about creating filters
to prohibit content or to control how people use the internet. Thus, some
elements of the software layer can be easily standardised in accordance
with technical needs, while others are very controversial.
The consensual model will also not work in the content layer of digital
infrastructures, considering the infinite variety of interests we may have
in any topic. The most probable is that the current trend becomes more
established through time: regulations will be defined by each country,
which is going to align these rules with the national legal framework. This
fragmentation is a source of conflict with global corporations that provide
services and products as part of digital infrastructures.
For instance, cloud computing services are affected by local legislation.
Technically speaking, data could be moved to any place on the planet
(or in space). From a technical perspective, data allocation in the cloud
should follow only the capacity of the servers and the proximity of servers
considering the business variables. However, national laws protect the
privacy of citizens and corporations in different ways. For this reason, cloud
service providers need to adapt their offering to each client considering
the local legislation for that particular client. This means that the same
cloud service provider is offering many different services, depending on the
country where each customer is located. We can see then that local laws
that affect only the content layer have an impact on the hardware aspect,
for instance, defining the level of investment in data centres in each country.
This demonstrates again the strength of the hourglass architecture model,
which shows that each layer influences the others. The same logic applies to
the governance and regulation of digital infrastructures.
A new area of concern from a governance perspective is the use of more
powerful AI tools. The more advanced forms of AI (which use machine
learning and deep learning) create complex mathematical algorithms based
on massive datasets (big data). The bigger the datasets, the more reliable
the algorithms will be in detecting and defining patterns in the data.
However, the problem comes that even the more skilful computer scientists
may not be able to explain the mathematics behind the algorithms. In a
regular piece of software, programmers and developers code the algorithms,
telling the machine how to execute a process. For instance, if A, then B. In AI
solutions, the algorithm is created by the software itself, looking for the best
possible match between the input and the desired output in historic data on
a similar domain (e.g. analysis of patterns and probabilities of events). With
the machine creating the algorithm, humans may not understand exactly
how the AI solution is concluding that B is the best output when asked about
A (to follow our example above).
What is the consequence of this? One could say simply that if the match is
good enough and if people are happy with the result, there is no need for
concern. We do not need to understand the calculation behind the results.
However, this means that we cannot guarantee the rules of good governance
are properly embedded in such an AI solution. If we cannot explain how the
machine has reached the answer B, then we cannot guarantee that answer B
is legitimate considering the laws and norms of a society. Two examples can
help us to understand the complexity of the topic.
179
Research has shown that some AI algorithms used in recruitment

of professionals have discriminated against women, particularly in
environments in which traditionally women have not been contracted
in the same way as men. Analysing these AI algorithms, it has been
discovered that because the datasets from previous recruitments have data
which reproduced the bias from recruiters, the AI tool has learnt the same
‘bias’ from the human recruiters. However, the problem now is bigger
than before. Why? If we know that humans may have biases, we may try
to reduce these biases through mechanisms such as having women also
on recruitment panels and being sure that the same number of women
and men are called for interviews during a recruitment process. But if we
think that the AI solution is neutral (the software does not have reasons
to have biases), then we may ignore these risks of biases and accept the
recommendation from the AI algorithm as being the best option available.
In this example, it becomes very difficult to guarantee good governance.
The second example is about the difficulty that AI algorithms have faced
to identify with precision the face of individuals who are not light skinned.
In this case, research has demonstrated that the problem is in the datasets,
which have a smaller representation of some ethical groups. If more
photos and videos are provided with a broader set of characteristics, the
AI solutions will learn to identify properly any individual (with the same
level of error for any ethnic group). Thus, the solution is to increase the
representation of dark-skinned groups in the databases. However, what
do we do while this problem is not corrected? Should we still allow AI
applications to be used to recognise faces? Again, the problem here is
of governance. If such a technology is used to give or not give access to
services, the implementation of a faulty technology means that some
ethnic groups will have difficulties in accessing services. Indeed, they
may be totally excluded from services if alternative channels do not exist.
Imagine even a worse situation, that such a piece of software is used by
the police to identify potential criminals in the street. As the margin of
error is higher for some ethnic groups, these groups would face a higher
risk of being wrongly identified as criminals. This example show that an
error in the AI algorithm may have serious implications for the lives of
people and the basic fairness of a society.
For these reasons, we need to discuss the governance of digital
technologies in society and embrace the idea of having more democratic
forums to discuss the development and application of technologies and the
speed in which such technologies are adopted. The objective here is not to
hinder the development of the industry, but to find means and resources
to guarantee that the development is coherent with our ethical values and
national and international legislations.
Another example of problematic governance happens with
cryptocurrencies, such as Bitcoin. Who is managing the access and
the savings of people in cryptocurrencies? Pieces of software, which
are distributed in peer-to-peer networks. In other words, nobody. It is
impossible for governments to control the operations inside peer-to-peer
networks similar to Bitcoin. When you go to the bank and you take a
significant amount of money from the branch, the government is informed
of this operation (in many countries). When you transfer money from your
account to another, the bank has this operation registered formally and
this information can be accessed by the government. In cryptocurrencies,
there is no control of how money moves around. For this reason, it is
easier for criminals to use cryptocurrencies for their operations. The
question then remains whether governments are going to allow or
180
support peer-to-peer cryptocurrencies. This is not about the technology

used by cryptocurrencies (blockchain), but the fact that the operations
in these peer-to-peer environments cannot be tracked and controlled.
Indeed, central banks may adopt similar technology gradually, but in an
environment controlled by the financial authority (not by peer-to-peer
networks). The Chinese government, as one example, decided in 2021 to
totally ban businesses with unregulated cryptocurrencies, such as Bitcoin.
It will not come as a surprise if other governments do the same in the
future.

This chapter focused on discussing the governance and regulation of
digital infrastructures. The chapter used the layered model (based on the
hourglass architecture) to discuss governance and regulation as layers
of digital infrastructures. It also used the metaphor of layers to divide
governance and regulation into three levels: the hardware, the code and
the content. The chapter focused mainly on the governance and regulation
of the internet, as the main digital infrastructure discussed in this subject
guide. In addition, the chapter presented some of the discussions related
to the other digital infrastructures. From a broad perspective, the chapter
worked to show that each governance and regulation layer affects
the others and that there are many conflicts of interest related to the
governance and regulation of digital infrastructures. Some examples are
provided about AI and cryptocurrencies.

• explain the relevance of governance and regulations in the way digital
infrastructures operate
• discuss the application of the hourglass architecture to the governance
layers of digital infrastructures
• outline the key arguments in some polemic discussions on governance
and regulations of digital infrastructures
• describe the relationship between the governance and regulation
aspects of the different layers (hardware, software and content) of

1. List and describe the key internet governance bodies. Discuss why
there are strong arguments about who controls these bodies.
2. Explain how the layered model may be applied to the discussion on
governance and regulation of digital infrastructures.
3. Discuss alternative models for the governance and regulation of
digital infrastructures (for instance, centralised models controlled
by governments, decentralised models coordinated by international
stakeholders, industry-drive models etc.).
181
Notes
182
Chapter 16: Issues of net neutrality and national borders
Chapter 16: Issues of net neutrality and

national borders
Chapter outline
This chapter discusses the topic of net neutrality and the difficulties of
thinking about principles for the whole internet as laws and regulations
are defined by national bodies. It takes the perspective that net neutrality
favours users and society as a whole, protecting citizens and democracies
against abuses from any sort of intermediary in internet communication.
In particular, the chapter presents the argument of empowerment, using
net neutrality as an instrument to empower society against abuses from
service providers and telecommunication companies. Some arguments
against net neutrality are also discussed. However, we take a perspective
here in defence of net neutrality.
16.1 Introduction
Net neutrality is the term used to define a particular form of distributing
information through the internet. In this model, ISPs, telecommunication
companies and governments (or any other third party between the content
provider and the user, or the sender of a message and the receiver) are
supposed to treat all pieces of information in the same way during the
transmission through the internet, neither discriminating against them in
terms of speed and priority of transmission, nor charging for any sort of
content in a different way, nor censoring or blocking any sort of content
that two parties wish to share. The same principle used for information
(content) should be applied to users, websites, platforms, applications,
access devices and forms of communication, which should not be
discriminated or charged differently. Net neutrality should guide the whole
internet traffic at any time and should be protected by legislation and
regulations, as proposed by those who favour this perspective.
This model is similar to that applied to telephone, water, gas and
electricity services – all regulated as public utilities. Electricity and water
go through the cables and pipes and the end user decides how to use it
in the household. Nobody but the end user decides how electricity and
water is used. There is no difference in fees for using the energy for the
computer or for the refrigerator, or the water for cooking or showering.
The same principle applied to the internet means that all pieces of
information travel without restriction through the network. Internet
traffic should be regulated as any other public utility. The end user is
responsible for deciding which pieces of information are required. Some
want to spend time on videos, others on news or emails. It is their choice.
Cable companies should be just carriers, without any power to interfere
in the communication between the user and the content provider (except
for cases where there is a court order with a particular objective). For
instance, if a peer-to-peer service is illegal in one country, it would be
legitimate to block this content. The decision would be based on laws and
regulation, not on private agreements with service providers.
Net neutrality works together with other internet principles to deliver
a service which is democratic and innovative. The internet works with
open standards to allow anyone to get inside the network. All the rules
of the internet are transparent. There is no censorship of what content
183
can be published (although people are responsible for the content they
publish and illegal content can be punished locally in accordance with
national legislation). With net neutrality, any two parts can interact on
the internet without interference of a third party. Without net neutrality, a
third party (ISPs, governments, mobile providers etc.) can interfere in the
interactions, reducing our degrees of freedom.

• Introduce the concept of net neutrality.
• Explain the advantages of net neutrality.
• Present the risks of not having net neutrality.
• Explore the conflict between the attempt to having common principles
for the internet and a regulatory framework which is defined within
national borders.

• outline key ideas related to the debate on net neutrality
• differentiate between two perspectives: empowerment versus market
forces
• elaborate examples of situations with net neutrality and without net
neutrality
• discuss the difficulties in establishing universal internet principles
considering the differences in legislations and regulations in each
country.

Net neutrality: https://en.wikipedia.org/wiki/Net_neutrality
US president Barack Obama’s arguments favouring net neutrality:
https://obamawhitehouse.archives.gov/net-neutrality
Internet Society’s view on net neutrality: http://www.internetsociety.org/
policybriefs/networkneutrality
Internet backbone: https://en.wikipedia.org/wiki/internet_backbone

Net neutrality in the USA: https://en.wikipedia.org/wiki/Net_neutrality_in_
the_United_States
Net neutrality in India: https://en.wikipedia.org/wiki/Net_neutrality_in_India
Net neutrality in Brazil: https://en.wikipedia.org/wiki/Brazilian_Civil_Rights_
Framework_for_the_internet

This chapter introduces the concept of net neutrality, comparing internet
services with public utilities. The argument is in favour of net neutrality
as opposed to a more laissez-faire attitude of letting market forces decide.
The chapter presents counter-arguments against key ideas used by
those who oppose net neutrality. The chapter admits there are areas of
ambiguity that society needs to discuss to establish priorities. However,
these limitations are not an argument against guaranteeing and regulating
net neutrality; rather, it shows that as societies we need to define the role
of the internet in our lives for building a more democratic and accessible
network in the future for the benefit of all.
184

16.2.1 The relevance of net neutrality
Net neutrality is a fundamental principle of the internet to guarantee
equality among users and content providers, transparency on the process
of transmitting data on the internet and an environment which favours
innovation. One of the ideals of the internet is that the network is open,
meaning that internet resources should be available to any person,
company, or organisation (including governments). Anyone should be able
to use the internet in the same conditions of others with similar resources.
Naturally, the information distributed by broadband has more speed than
the information distributed in a dial-up connection, but this difference
comes from the features of the technical infrastructure, rather than from
someone (a gatekeeper) deciding that one user should get more speed on
the internet than others.
There are many ways of breaking the net neutrality principle and,
unfortunately, we have seen many abuses of power for companies and
governments that become the gatekeepers between content providers and
users. For instance, a service provider can block or reduce the speed of a
particular communication protocol, prohibiting peer-to-peer file sharing or
videos or emails in its network. In this case, a particular sort of protocol is
penalised. Let’s say that this service provider realises people spend more
time watching videos than using the internet for other purposes. It could
then slow down the speed of transmission of videos and force all content
providers of videos to pay more to regain the normal speed so that they
can reach their audience.
A second way of breaking the net neutrality principle involves blocking
certain IP addresses or treating them differently; a powerful method of
censorship used in less democratic countries. ISPs could use this model to
impose fees on content providers that demand more bandwidth or which
have more users. Through this model, an ISP can decide that Google
Search is the channel for Google to make money with advertisements,
so either Google pays some of its profits to the ISP or the ISP blocks
Google Search. In this case, if Google does not agree to pay more to the
gatekeeper, the users of this service are not going to have good access
to Google Search, in spite of their preference for using the service. Less
democratic governments use this approach to filter websites they consider
to be opposed to their political interests, either filtering the internet
backbone (principal data routes between the largest and more important
computers networks connected to the internet) to the whole country
(national firewalls) or imposing on national ISPs the cost of filtering the
content in their networks. Sometimes, governments force ISPs to verify
whether users are trying to get information on prohibited or political
topics through search mechanisms, as a way of controlling the online
activities of citizens.
Alternatively, the ISP can ask a fee for giving priority to the content of
a particular provider. Let’s say that one big e-commerce service buys
priority in all ISPs in a country, reducing the speed of access to any other
e-commerce website. In this case, this e-commerce service would mainly
control digital e-commerce in the country, as the other companies would
not be able to compete on the same networks. In this case, the user would
be able to get quick access to one e-commerce service and very slow access
to others, in practice reducing their options and distorting the competition
in the market. People will not wait indefinitely for a site to be accessible,
while there are other sites that are available immediately.
185
Activity 16.1: Defining net neutrality

Define the concept of net neutrality. Find examples of how companies and governments
have been avoiding net neutrality around the world. What are the consequences of these
attitudes?
16.2.2 Arguments favouring net neutrality

Some ISPs and telecommunication companies argue that they have
invested lots of money in building their networks, so the way they
distribute content in their infrastructure is their own businesses. From this
perspective, just as a supermarket decides which products go on its shelves
and the price of each product, the ISPs have the right to decide which
pieces of information are distributed (and the price content providers and
users pay to get the information). If a user is not happy with an ISP, just
contract another service provider, in the same way a person can change
supermarkets. The invisible hand of the market would regulate the access
to information.
However, we need to recall here the way the internet works though, to
argue that, in this case, there is no such competition in the market. It is a
network of networks, connected by the same protocol. For one person in
China to access a website in the USA, they need an ISP which is connected
to all other networks until the moment the information request reaches
the servers of the content provider in the USA. It is a long way to go, from
the Chinese service provider to the internet backbones, till reaching the
US service provider which gives access to the US content provider. In this
model, we see that the final user and the content provider depend on the
local ISPs and on the telecommunication infrastructures of many countries
to be able to communicate with each other.
Detractors of net neutrality argue that the free market will resolve the
issue, as people will be able to opt for alternative ISPs, and they will
choose the provider that is the best gatekeeper of their interests. But
this does not happen in practice. In any locality, we do not have many
service providers to choose from, just as we do not have too many
electricity, telephone, gas and water suppliers. As investing in building
the infrastructure for local networks is expensive, in many places around
the world a household will have one or at the best two service providers.
Even if some places have five or six service providers, the argument
favouring market competition is fundamentally flawed because it ignores
the discussion on who has the power to control the internet. Today, users
navigate the internet without gatekeepers. They pay for ISP access, but
they decide what to access. Why would a user be interested in having less
freedom?
For those who defend net neutrality, the actual relevant argument is
that societies cannot allow some economic actors to have such power of
defining the internet traffic. Digital communication is too important a right
for societies to delegate to market forces and private companies the power
to make decisions in the name of all. The argument goes that societies
should guarantee by law and regulation the right to net neutrality in order
to avoid situations in which gatekeepers maximise their profits instead of
maximising benefits for users and society. The public good should be above
the interest of private companies.
It is in this context that many defend the idea that net neutrality should
be regulated by law, to protect the interests of society, blocking ISPs,
telecommunication companies and any other party from getting control of
186
the internet. Access to information is a valuable asset to societies, which

requires governments to create proper legal frameworks to protect the
interests of citizens against abuse of companies, in the same way that
other essential services are regulated, such as education, healthcare and
the provision of utilities (communication, water, gas and electricity).
Otherwise, ISPs and telecommunication companies will become the actual
gatekeepers of the internet, with power to define who is going to access
any content in their networks, becoming the real controllers of a digital
infrastructure which is supposed to not have any authority in control.
From this perspective, the lack of regulation goes against the principle of
net neutrality and allows abuses to happen.
Those who are against net neutrality regulation, their arguments focus
on the competitive forces of the market. Leave it to the market to decide
who will be the gatekeepers of the internet. They also argue that having
regulations imposing net neutrality will reduce innovation in network
infrastructure: if service providers cannot make as much money as they
would like, they would not build their networks. This argument has a
logical flaw though: the service provider may charge the final user in the
way they wish – indeed, there are many different fees for people with
different needs in terms of quantity and quality of data transmission.
The price fixed by the service provider is the best way of having market
mechanisms: the provider will find the best price to get more users and
maximise the profits. This is a fair market based on price mechanisms only,
without making the service provider a gatekeeper of internet traffic. If one
pays to have the service, one accepts the price. The profit will come from
the revenues and not from asking for more fees from content providers.
From the perspective of net neutrality, it is legitimate for a company to
increase prices to obtain resources for more investment, as this does not
bias the access to content.
If there are other obstacles to investing in internet infrastructures, more
should be discussed in terms of taxation and avoiding local red tape,
to protect investments and innovation through other means. In some
situations, governments may need to provide the internet infrastructure
to speed up the development of a country. In this situation, the focus
would be on the investment in the infrastructure, without empowering the
gatekeepers to decide on the transmission of content. In countries with less
well developed telecommunication companies and ISPs, such incentives
may be necessary to ensure the population and organisations have access
to information.
Indeed, there are strong arguments that net neutrality favours innovation,
reducing the entry barriers of those who want to challenge incumbents.
Today, if an innovator offers a better service, anyone can choose to use
this better option. Think about Facebook. In the beginning it was only a
small social network challenging incumbents such as MySpace and Orkut.
Today, Facebook is a service with more than 2.9 billion users because
people have decided it offers a better service than others. Facebook would
not be possible in a world in which incumbents could make deals with
ISPs. Imagine a situation in which MySpace could pay ISPs to slow down
the speed of access to Facebook, protecting its position against the new
competitor, which would not have resources to pay for a better access.
Indeed, the internet is full of innovations which may not be known by
most, but which are good enough for their users and thus reward the
content provider. Mobile app services are good examples of this explosion
of innovations, with millions of different options for users to choose freely.
Without net neutrality, the internet would become a place for those who
187
have the means to get better deals, instead of being the place for better
innovations from the perspective of users.
There are also reasonable arguments that different types of data should be
treated differently in accordance with the order of priorities of a society.
For instance, a society may decide that emails and voice data should have
priority over videos and that videos should have priority over games. This
would imply that every time a network is very busy, games and videos
would need to wait till the emails and voice messages been transmitted.
This is a sort of discrimination by type of data, which attend the needs of
users and service providers to maximise the benefits for society, without
empowering the service providers as gatekeepers. We can use a metaphor
here: ambulances and firefighters have priority in public roads and all
agree they should have priority for the good of society. A similar rule on
data priority could be defined by regulation, or by the contract between
the service provider and the final user (if the priority would be decided
by the user, not by the ISP). But this would not allow a service provider
to charge more for data coming from Google than the data coming from
Amazon, to give an example. In this model, net neutrality would not work
as it works today, but the freedom of accessing content from different
providers of same kind of information would be the same.
Activity 16.2: Understanding both sides of the argument

Research key arguments for having legislation protecting the net neutrality and
arguments against protecting net neutrality. Compare both arguments. Who is going to
benefit more from having legislation protecting net neutrality?
16.2.3 Examples of net neutrality benefits

The best way to show the benefit of net neutrality is to discuss some
examples of potential scenarios if we do not create clear legislation
protecting the interests of people. Let’s suppose that ISPs may decide
which information packages are going to be sent quickly and which ones
can be left behind or even not transmitted at all in their networks. In
such a scenario, an ISP could make an agreement with Company X that
its search engine would work quickly in the customers’ browsers, while
the Google engine works slowly. In this hypothetical situation of traffic
shaping, the users of this ISP would be faced with two options: either
using Company X’s service which is quick, or paying the price (in terms
of time and delay) for using Google. Even worse, the ISP could decide
that Google Search would not be accessed in its network. The user could
also change the ISP, but the problem would be the same: the new ISP, not
the users, will also have the power to make decisions on which content,
websites and services would have priority in its network. In addition, the
user could face a problem in finding another local ISP.
In this scenario, the ISP would be able to gain twice for being the
communication channel between the internet backbone and the user:
firstly, getting a fee from the user (that fee you pay monthly to have
internet at home); and secondly, getting a fee from the content provider
to get its services delivered. In this case, the ISP would be the one who
gained most from not having net neutrality, as content providers will be
forced to pay them a fee to get their content delivered. The profits of
content providers would be reduced substantially if the power of decision
on which goes on the internet were transferred to ISPs. Innovators could
be blocked, as incumbents could pay for ISPs to block new competitors.
188
The user would lose immensely. By contrast, with net neutrality, we have
the freedom to decide which content we want to access and the capacity
of reaching innovations and choosing when an innovation is good for us.
Without net neutrality, the ISP would somehow decide what the users
are going to access. Once you realise that Google Search does not work
anymore in your computer (or it works very slowly), you would use the
search in other interfaces which have business deal with your ISP.
The same logic of filtering can be done because of other reasons. Let’s
say that the owner of a powerful ISP has a political view on an issue, or
is supporting a candidate in an election (or even is the candidate in an
election). The company would be able to filter the content in its network,
favouring the good news about the candidate or political perspective
and eliminating the view of others, or the bad news about its candidate.
Indeed, many news companies have done this in the past and present,
filtering and curating content, pretending to be neutral to convince readers
of a particular perspective. The internet breaks the monopoly of truth of
too power communication companies: anyone can express their ideas and
people can share ideas freely. Once we allow ISPs, telecommunication
companies and governments to filter content, we can no longer be sure
of the quality of the information. We would also not be able to guarantee
we have access to a broad discussion of different perspectives. It would be
like parents who use the parental control software to decide which content
their children can access. The ISP can use similar filters to control which
content we can access if net neutrality is not protected.
Let’s take an example of discrimination by type of user. Suppose a big
corporation makes agreements with all ISPs in its neighbourhood to have
high priority in the delivery of internet services. This would mean that any
time the network is very busy, the company would have priority to receive
its information, with the implications that the internet would be slower
for all other users in the same area. In this case again, the ISP would
have the benefit of getting more fees for selling special services to the
big corporation, the corporation itself would benefit from having priority
in receiving information and the other users around would have worse
service quality, although they are paying a fee for the services. Probably,
the users would even not know that they are receiving worse services,
despite paying their subscriptions, because the service providers have
decided to make extra money selling priorities to the big corporation. How
often have you thought your internet connection slow compared it with
the quality of the services your provider has promised?
With net neutrality, we the users decide which content is good for us.
It could be otherwise if net neutrality is lost. At the extreme, each ISP
could decide which websites would be available or not in its network.
Indeed, we know very much how this would work, because we have
seen it before. Without net neutrality, the internet would repeat the old
model of traditional broadcast systems, in which the broadcaster curates
the content that audience can access. This traditional broadcast system
empowers the broadcaster and limits the options available to the audience.
In many countries, broadcasters have monopolistic positions in markets
as the only providers of communication and information services – such
as radio and television channels. If net neutrality is not present, the ISPs
and governments would be able to become the curators of content for the
users, with unthinkable damages to our freedom of choice. Just compare
the difference of access to content you have via television and YouTube.
189
Activity 16.3: Comparing the internet with broadcasters

Make a list of methos of access to content on the internet in a democratic country which
respects net neutrality. Make a second list of characteristics of how a person can access
content from broadcasters in the same country (such as radio, television channels and
cable TV), without using the internet. Compare both. Which model most empowers the
audience? Why?
Activity 16.4: Being a consultant

Suppose you are a consultant on digital infrastructures. You are providing consultancy to
a government which does not accept open opposition to its political views in any sort of
media. What sort of advice would you give to this government on net neutrality?
16.2.4 Universal principles versus national borders

From the discussion above, it is implied that any attempt to protect or
reject net neutrality depends on national legislation and regulations. As
discussed in Chapter 15, it is a big challenge to try to define governance
and regulations for digital infrastructures which are, technically at least,
global structures with pieces which are regulated by national governments.
When the matter is really technical, such as a communication protocol,
it is easier to get an agreement (although it may take a long time to
reach consensus), considering all parties are interested in keeping the
communication flow. Even countries which adopt firewalls to block some
sorts of internet content keep using the same communication protocols,
because overall they do not want to isolate themselves from the internet.
However, when the situation is political or affects business interests,
more conflicts and disputes arise. In our example in this chapter, of
net neutrality, it is clear that we have stakeholders with very different
perspectives and that national legislations can either empower the people
and society as a whole or business corporations and governments.
From the business perspective, powerful industries and corporations
have interests in influencing the internet data flow. These stakeholders
may lobby local governments to get their interests reflected in local
legislation, or to block legislation in favour of net neutrality. Civil society
organisations, academics, researchers, consultants are also lobbying for
their own view on the matter. It is interesting also to observe that similar
stakeholders have different views depending on which national legislation
they are trying to influence. Companies have different views in accordance
with their share of a particular market. National governments also are
going to make decisions considering who those defending the idea of net
neutrality are. For instance, major internet corporations such as Google
and Facebook face challenges in countries in which legislators do not like
the fact that these giants are US companies. In political games like that,
the interests of the users and society may be overridden by the interests of
lobbyists and politicians.
Internet censorship clearly goes against the principle of net neutrality
but is not strictly the same as net neutrality. Net neutrality itself is about
how the IP layer allocated bandwidth on a first-come first-served basis.
It is fundamental to the protocol layer. However, if net neutrality at the
‘packet’ level is lost, then all kinds of influences on data flows can find a
way in. Certainly net neutrality alone is not a way of imposing freedom
of information. For example, governments and ISPs can use firewalls
190
to block services they consider to be dangerous to their interests. Still,

the discussion on net neutrality is part of a broader debate on openness,
freedom of information, democracy and forms of collective and consensual
decision making. We certainly should not be fooled into believing that in
democratic countries and representative democracies the interests of people
are necessarily going to be fully respected. In all democratic countries,
powerful economic and political groups exert influence through lobbying,
to get the benefits they want in the name of benefiting the whole society.
For this reason, citizens perhaps also need to organise themselves to be sure
that their interests are better represented in this discussion.
Activity 16.5: Comparing legislations

The best way to learn about net neutrality is to understand the discussion on related
pieces of legislation. Research the legislation of your country to know how net neutrality
has been protected. Do not be surprised if nothing is written on this matter. Indeed, many
corporations against net neutrality lobby for not having any legislation on the matter, to
allow them to do whatever they want. If your country does not have a legislation, look at
the discussions held by parliament or other legislative bodies on this topic. What are your
conclusions about net neutrality in your country?

This chapter discussed the principle of net neutrality. The chapter gave
many examples of situations in which net neutrality brings more benefits for
users and society. The key argument used here is of empowerment of society
versus a market perspective of allowing service providers and intermediary
of internet communications to become gatekeepers of the content shared
through the network. There are other arguments used by those against net
neutrality, which have been explored through the exercises. The chapter
ended by explaining the difficulty of building universal principles in a
digital infrastructure which by nature is fragmented across countries.
There are thus limitations in trying to reach any sort of universal principle
considering the context of different national legislations and regulations
applying to the internet.

• outline key ideas related to the debate on net neutrality
• differentiate between two perspectives: empowerment versus market
forces
• elaborate examples of situations with net neutrality and without net
neutrality
• discuss the difficulties in establishing universal internet principles
considering the differences in legislations and regulations in each
country.
191

1. Define net neutrality in your own words, comparing this definition
with the idea of public utilities.
2. Contrast the arguments that the service providers should be free
to define the terms of transmission of internet content with the
arguments that net neutrality is about empowering users and society.
3. Discuss the implications of having national legislations and regulations
applied to the same global digital infrastructure of the internet.
192
Chapter 17: Digital business platforms and digital infrastructures
Chapter 17: Digital business platforms

and digital infrastructures
Chapter outline
This chapter aims to develop the concept of digital business platforms.
The objective is to link previous discussion on digital infrastructures,
developing a conceptual framework based on the concept of platforms.
The chapter analyses examples of platforms in digital industries,
explaining the products, services and technologies they offer to enable
third party businesses. The perspective developed in this chapter relates
platforms to the concept of digital infrastructures, as they have many
similarities in terms of developing themselves in layers and enabling other
business to be built on the top of their existing layers.
17.1 Introduction
In previous chapters we have discussed digital infrastructures, focusing
mainly on the examples of the internet, the WWW, cloud computing,
mobile technology and social media. We have argued across this subject
guide that digital infrastructures grow in layers and that they are
fundamental for business innovation in contemporary economies, as they
enable many other businesses.
In this chapter, we conceptualise digital business platforms, which have
many similarities with digital infrastructures. In the same fashion as
digital infrastructures, digital business platforms evolve through time in
layers, enabling other businesses, which develop their business models,
products and services on top of these platforms. The key difference, from
the perspective of this subject guide, is that digital business platforms are
mainly information systems specialised in particular goods and services,
which run on the top of broader digital infrastructures.
For instance, the internet is a digital infrastructure, providing the
environment for all sorts of digital products and services we have
access to online. It is a very broad resource. The same applies when we
conceptualise social media as a digital infrastructure. We are talking about
hundreds of different applications which have enough similarities to be
aggregated into a category, but at the same time are very diverse. These are
very generic infrastructures, which bring together a large variety of layers.
Digital business platforms are more specialised, enabling a particular
segment of digital products and services, contributing to a well-defined
value chain or business exchange among economic players. For instance,
Facebook is a service which has become a fundamental layer of the
social media infrastructure. However, Facebook is also a digital business
platform, as it provides the infrastructure for other businesses to be
built on top of its service. Through APIs, Facebook allows companies to
build complementary apps such as games, to be used through Facebook.
Similarly, Amazon is one of the most important e-commerce layers
of the internet and mobile technology infrastructures and one of the
most important cloud service providers, being thus a layer of the cloud
computing infrastructure. However, Amazon is also a digital business
platform, allowing companies and developers to create apps and business
solutions on top of its services, increasing their capacity of reaching
193
customers (thus adding value to the business chains which connect those
companies and their customers).
From this introduction, we can see that digital business platforms
are closely related to digital infrastructures. We use the term digital
infrastructures to conceptualise the broader assemblages of information
systems (hardware, software, people, processes and governance layers),
which enable a very generic and large number of applications. The term
‘digital business platform’ refers specifically to the intermediary layer
which enables a particular set of more specialised products and services.
Digital business platforms are thus layers which operate on top of broader
digital infrastructures, enabling other layers to emerge.

• Introduce the concept of digital business platforms.
• Relate digital business platforms to digital infrastructures.
• Show the link between digital business platforms and innovation.
• Provide examples of digital business platforms.

• explain the difference between digital business platforms and digital
infrastructures
• describe the links between digital business platforms and digital
infrastructures
• identify key digital business platforms
• describe the link between digital business platforms and innovation.

Gawer, A. and M.A. Cusumano ‘Industry platforms and ecosystem innovation’,
Journal of Product Innovation Management 31(3) 2013, pp.417–33.

Gawer, A. ‘Bridging differing perspectives on technological platforms: toward
an integrative framework’, Research Policy 43 2014, pp.1,239–49.
Gawer, A. and M.A. Cusumano ‘How companies become platform leaders’, MIT
Sloan Management Review 49(2) 2008, pp.28–35.
Iyer, B. and T.H. Davenport ‘Reverse engineering Google’s innovation machine’,
Harvard Business Review 86(4) 2008, pp.58–68.
Parker, G.G., M.W. Van Alstyne and S.P. Choudary Platform revolution:
how networked markets are transforming the economy and how to
make them work for you. (New York: W.W. Norton & Company, 2016).
[ISBN 9780393249132]

This chapter presents the concept of platforms in general and more
specifically focuses on those which are related to the digital industries –
the digital business platforms. It draws upon the work of Anabelle Gawer
and Michael A. Cusumano, adapting their focus to the objectives of this
subject guide. We focus here on the industry-wide or public platforms,
rather than on company-specific ones. The chapter shows that digital
business platforms are layers of broader digital infrastructures and that
they have a role in creating ecosystems of innovation. After presenting the
conceptual framework, the chapter applies the concept of digital business
194
platforms to demonstrate how a number of companies specialising in

digital goods and services have become enablers of other businesses. The
examples are not an exhaustive analysis of each platform, but an effort to
show a variety of cases which may better inform you on how to apply the
framework in other contexts.

17.2.1 Digital business platforms
The concept of platforms is used in many industries, but in the last decade
it has become more pervasive when discussing the nature of businesses
which rely substantially on digital technologies to offer services to others.
In this subject guide, we refer to this group of companies as digital business
platforms (DBP), as they specialise in enabling or supporting digital
business. Some authors may call them technological platforms; however,
here we want to emphasise the digital business aspect. Companies like
Alibaba, Microsoft, Apple and Google are examples of large digital business
platforms that sustain ecosystems of innovation, putting together hardware
and software layers, connecting to different devices, creating networks
of users – individuals and organisations – as creators/developers and as
producers and consumers. There are of course many smaller and more
focused digital business platforms too, often focused on a specific industry
sector like travel, finance and investment or transport.
To be a digital business platform is to enable others to do digital
business, in the most general sense. A DPB can be seen as an
intermediate application layer in the layered model of infrastructure.
DBPs, as infrastructures, conforming to the six characteristics of digital
infrastructure in Figure 2.1. A DBP offers digital services and products
which are used by (many) others, helping them enable and further
develop their individual businesses. A DBP will provide necessary and
needed services and, as a generative infrastructure, enable innovation in
terms of new products or services and support for new business models. A
successful platform needs to have a critical mass of users to bring positive
network effects to all users. For this reason, in their early phase in business
DBP often subsidise some users to attract them to join the platform and
help them achieve scale.
Drawing upon Gawer and Cusumano (2013), we can define platforms
as ‘products, services and technologies which are organised as
interdependent elements into a common structure and sets of processes,
which allow a company to enable others (individuals and organisations) to
develop their own products, services and technologies to be offered to the
market or used by a firm or an individual.’
Starting from this definition, we might in some cases identify any
particular platform with associated services and technologies, such as
Microsoft Windows or the Linux operating system. We can also think
about platforms which facilitate many products and associated services
and technologies, enabling a wide range of business functions and with
potential for innovation (generativity). For example, the Google platform
includes search, email, maps, browser, mobile operating systems, cloud
services, advertising, media and video etc).
From an economic perspective, digital business platforms create value by
enabling transactions between different parties, some being suppliers of
products and services and others being customers/users. Both parties are
able to interact productively because of the service offered by the DBP.
195
Another way to see this is to say that digital business platform is a market
place; as markets they benefit from network externalities and each party
(buyers or sellers) gains more benefits the bigger the numbers of the other
party. For instance, from a user perspective, the more developers of apps
in a mobile platform, the better. From the developer perspective, the more
users in a mobile platform, the better. From the platform’s perspective,
the more users and developers, the better so that economies of scale can
combine with network externalities. Because of network effects, there is a
self-reinforcing mechanism encouraging platforms to pursue growth and
we may have situations in which the winning platform takes the whole
market (or almost all of it). See, for instance, the comfortable position of
Facebook in the market of digital social networking, of Google in search
and advertising and of Android as a mobile operating system. In cloud
services, a large and important type of DBP there is in contrast, a highly
competitive bunch of big companies and less a sense of ‘winner take all’.
The emergence of digital business platforms has consequences for whole
industries if and when a digital business platform becomes a key player,
with other economic actors losing some of their relevance and market
power. The more powerful a DBP and the more other businesses become
dependent on it – locked in - there is an increasing chance that a bigger
share of value produced is transferred to the owners of the platform. We
see this in recent decades with the increasing revenues of companies
such as Apple, Amazon, Alibaba, Google, Microsoft or Facebook, to name
just a few. The more that economic agents become dependent on digital
technologies, the more value these platforms are going to have.
Digital business platform leaders, as powerful intermediaries, need to
balance competition and collaboration, not trying to squeeze the profit
margins of their users too hard. Here the term ecosystem is a good
metaphor: despite their powerful position, platform leaders need to keep
fostering the businesses which run on their platforms. They thrive when
their users thrive. This co-dependency between the platform and the
businesses it supports mean that the design evolution of a DBP needs
reflect this. For example, by supporting collaboration within the ecosystem
as technology evolves and recognising that changes in the platform are
going to affect all businesses related to it. A successful DBP cannot (or
should not) make such decisions alone without considering the impact on
the ecosystem.
Activity 17.1: Applying the concept of digital business platforms

In this section, we defined the concept of digital business platform (DBP). Research online
how Facebook has become a key digital business platform in the advertising market
and beyond. Particularly, find how Facebook offers a differentiated service to match
advertisers and consumers. Think about the unique features of the Facebook platform,
from its Pages and Groups facilities to its algorithm which chooses the ads to be shown
to each user. What are the key differences between the advertising services offered by
Facebook and the ones offered by Google (AdWord and AdSense)?
17.2.2 Enabling innovation

Digital business platforms can enable innovation in products and services.
Thy do this by adopting a modular architecture which allows the creation
and recombination of blocks. As we have seen in earlier chapters, this is
a basic characteristic of digital infrastructures and favours innovation by
allowing complex systems to be broken into manageable components,
which interact with each other through standardised interfaces within
a defined architecture. Innovation can come from the recombination
196
of modules in different formats, sometimes guided for a new business

model (see Chapter 18). Modularity also facilitates innovation in so far
as developers do not need to know the whole system but can focus on
specifics and limited areas of change and novelty.
By having a modular architecture, components of a DBP can be used and
reused across the many products, services and technologies, bringing
economies of scope for the platform itself and all who participate in it. The
economy of scope means that the cost of producing new combinations of
components is reduced relative to producing the same product or service
from scratch outside of the platform. For instance, a hotel owner can
benefit from Google Maps service, adding to the hotel webpage a map
with its location. The alternative of finding a suitable map or building
a map would be certainly more expensive. Similarly, using Amazon as
a platform for e-commerce is a cheaper solution for small and medium
businesses than having their own website, with proper security and
payment procedures.
The interfaces of a digital business platform are in general open to allow
the creation of complements and innovative plug-ins. Indeed, the more
innovation is created on top of a digital business platform (new layers,
new apps, new business models), the greater the value of the platform
for all economic actors who participate in the ecosystem. The level of
openness of digital business platforms to third parties varies, affecting the
way these parties are going to be related to the platform. The openness
depends on how much access to information is offered to the third party,
as well as the rules of the interaction and the cost for participating in the
platform. In this way different architectures and institutional arrangements
will foster more or less innovation.
It is also necessary to recognise that there may be competition between
the platform and those who develop solutions to run on the platform. In
this situation, the platform owner may not benefit from innovations that
directly compete with the platform’s services. An interesting example of
platform–platform competition is when Facebook created an interface to
operate on Android platforms. In trying to become the ‘home screen’ for
users of Android (owned by Google), Facebook tried to hijack the platform
for its own benefit, taking users away from the Google ecosystem and into
the Facebook one
In any case, the level of innovation probably should be curated by the
platform owner to make sure that the whole ecosystem keeps benefiting
from it. Too many components may make the system too complex, creating
a barrier to more innovation. Third parties might start to avoid creating
new solutions for a platform that already has too many overlapping
options. It may be necessary to select from the innovations those that
better contribute to the whole system. This happens, for instance, in open-
source software platforms, when the community of developers decide
which innovations are going to be incorporated to the system, rejecting
innovations which are not contributing to the improvement of the whole.
Thus a platform leader needs to keeps a level of control of the architecture
of the whole system, avoiding the shortcomings of too much complexity.
On the other hand, as organisations and individuals aggregate around
a few huge digital business platforms (because of the positive network
effects), some digital business platforms have tried to centralise an
increasing amount of power, enabling them to create barriers for other
companies entering the market with competing solutions. This may
reduce innovation in the long term, going against one of the key benefits
of having platforms. The question thus is how to keep a balance between
197
the interests of incumbent platforms vis-à-vis the market’s needs for

innovation and the interests of all other suppliers and customers who
operate on the platforms.
Activity 17.2: Innovation in digital business platforms

Drawing upon the idea that digital business platforms foster innovation, research online
how many apps you can find available in three different mobile operating systems
(Android, iOS and Windows). Analyse how these platforms enable innovation through
open APIs and the processes they use to manage innovation such as the way they accept
or reject apps, how they display key apps and how they rate apps. How easy is it for users
to find the apps they need?
17.2.3 Linking digital business platforms to digital infrastructures

In this section, we look at digital business platforms a specific type of
digital infrastructures and spell out some of the similarities between
the two.
The first similarity is related to the way digital business platforms gain
scalability, evolving through time, drawing upon initial designs, but
drifting in directions which have not been originally planned. Although
usually owned by a company, which controls the design of a core part
of a platform, the platforms we see emerge over time reflect the way
core components, organised in modular structures, are assembled, used
and recombined by users (individuals and companies, suppliers and
customers). In other words, they evolve reflecting the environment in
which they operate and the products, services and technologies which are
created there. Thus there is co-creation and co-design of digital business
platforms, in which the company which controls the platform adjusts its
design and architecture to gain scale, exploit linked innovations and so as
to become more attractive to more users.
The second similarity is about the development of layers. The development
of new products, services and technologies happens (in part) through
adding new layers and new nodes to complex networks of technology,
organisations and people. The ability to add new layers depends on the
degree of openness of a digital business platform, which requires clear
interface standards and governance structures. Digital business platforms
usually are designed as modular architectures, with core components and
the capacity of adding more modules in accordance with the needs of
users. In addition, users contribute to the creation of new elements, for
example as apps which run on the top of Android or Apple platform or
games on Facebook.
The third similarity is that both are strongly affected by network effects:
positive and negative externalities. On the one hand, the more participants
a digital business platform has, the more benefits for the whole network.
For instance, the fact that Facebook has a massive number of users is a
benefit for those advertising on the platform, as this increases the chances
of reaching potential customers. The network externality in this case
benefits advertisers and Facebook itself and perhaps the users too (do
you benefit from adverts?). On the other hand, the massive number of
participants may also bring negative externalities. For instance, too many
companies trying to sell social games on Facebook may make it harder
to convince a particular user to try a new game or to accept an invitation
of a friend to play a game. Excess of information can also become a
shortcoming for a digital business platform when too many participants
are connected.
198
17.2.4 Some examples of digital business platforms

This section discusses some examples of digital business platforms. It
does not aim to give a detailed account of each platform; rather it has the
objective of summarising key ideas about a few examples for the student
to investigate more in the exercises. Reading the cases, it is evident that
all these platforms depend on one or more digital infrastructures, from
the broad internet and the WWW to the more specialised ones: cloud
computing, mobile technology and social media.
Case 1: Google and Android

Google has today one of the most pervasive digital business platforms,
with a business model which allows the company successfully to offer
a wide range of services to end users. The main service offered by the
company is Google Search, which has the biggest market share in this
market. Since its earliest days, Google has aimed to become a major
digital business platform, offering a collection of complementary products,
services and technologies which are all integrated in the same structure
and aim to facilitate individuals and businesses to do business on Google’s
systems and generate data for Google at the same time. For instance, the
company offers email services (Gmail), the Chrome web brower, office
software for collaboration (packages with Docs, Sheets, Forms, Slides and
Sites), storage (Google Drive, a cloud service), video platform (YouTube),
blog interface (Blogger), calendar, video conferencing (Hangouts), Google
Maps (and Street View), Google Images, Google Books, Google Flights,
Google Shopping, services for advertising such as Google AdWords and
Google AdSense and many other services. Individuals and companies may
use these services in different ways in accordance with their needs. Some
services are free (especially for final users) and others are paid (especially
for corporate use of the platform).
Having realised the relevance of mobile infrastructures, Google took a
bold step to gain the biggest market share of mobile operating system
by launching Android, a piece of software based on the Linux kernel (an
open-source operating system). The Android operating system is targeted
at touchscreen mobile devices such as smartphones and tablets. Android
has been also adapted for television, cars, wrist watches (wearable
technology), notebooks, game consoles and digital cameras, among other
devices. Android is a key part of Google digital business platform and
has in itself become a major platform. Android has an open API, allowing
developers around of the world to produce a vast range of apps on the top
of Android and offer them through the Google Play Store.
Case 2: Apple and iOS

After struggling in the 1990s to redefine its identity in the PC market,
Apple has rebuilt itself to become one of the key digital business platforms
in the world. Apple has reinvented itself as a platform, migrating from
a more product-oriented strategy to a combination of products, services
and technology. Distinct from Google, Apple has its core business as the
sale of hardware – computers (Macintosh), mobile phones (iPhone) and
tablets (iPad), television sets (Apple TV), wearables (Apple Watch) and
music players (iPod), among other products, including accessories for
these products. Apple creates its own software, supported by proprietary
operating systems, OS X for Macintosh and iOS for Apple mobile devices.
On top of these hardware and software layers, they offer services such as
iTunes, App Store, Apple Pay (a propriety payment system), iCloud (cloud
services), Apple Mail and Apple Remote Desktop, among others.
199
Relying on the perceived quality of its products and a loyal customer base,
Apple has been able to expand the sales of both, hardware and services,
changing its role in the digital industries. An important move in this
direction was the creation of the service iTunes, associated at first with the
iPod product. The iPod device is now (2022) almost totally superseded
by mobile phones, but it was very innovative in its day and carried digital
music (and other files too). However, the real value added by the device
was the possibility of buying songs through the iTunes platform, which
uses cloud services to store and manage music, videos, television shows,
audiobooks, podcasts and movies rentals. iTunes content is available on
the other Apple devices too. The interesting point about iTunes is that it
opened the door to a new business perspective for Apple. The company
adopted a business model of intermediating the business between the
final consumer and the recording companies, offering a platform to sell
individual songs (instead of CDs).
Apple develops operating systems and software which operates only on its
hardware/platform, creating an environment in which the final user has
a combination of hardware, software and services sourced from Apple. In
order to leverage the level of innovation in its platform, Apple has opened
the APIs of the web browser Safari and of the iOS for mobile devices. In
doing this, Apple has benefited from thousands of developers around the
world who have developed software solutions (apps) to run on top of
Apple devices and software. As with Google and Android, Apple launched
the App Store for selling the apps created by these developers, thus on the
one hand organising and managing the offer of apps in its platform and
on the other hand making viable a market to match users and developers,
which would not be easy without Apple intermediation. And of course, the
company gets a percentage of the value of all sales through the App Store.
Case 3: Microsoft
Microsoft has been a key player in digital industries for many decades.
The company started offering operating systems and office packages for
PCs. The Windows operating system is the most important system on
desktop computers and smaller servers and supports the larger Microsoft
platform. In addition the company is big in games consoles (Xbox). The
company offers many software packages as services such as Word, Excel,
PowerPoint, Edge, Outlook email and associated services (such as calendar,
contact lists and management of tasks) and collaborative platforms such
as Skype, SharePoint and Teams. By virtue of these services that are widely
used in business, Microsoft is a very significant digital business platform.
There are few businesses or other organisations that do not use some of
these products and services offered by Microsoft although the company
has competitors in all segments. Microsoft has also increased the range
of service platform offers to include significant industry leading cloud
services – Microsoft Azure. And its own mobile Windows operating system.
And in a similar way to other DBPs, Microsoft has an open API to allow
developers to create new apps to its platform.
Activity 17.3: Understanding Microsoft cloud services

Research online the cloud services offered by Microsoft Azure. Check the service itself
and the business benefits they may bring to companies in terms of data management,
enterprise mobility, collaboration and security. Who are the principal competitor cloud
platforms for Microsoft?
200
Case 4: Amazon
Amazon started as an e-commerce business: selling books on the internet.
Over many years, Amazon operated at a loss, as the market was not yet
mature enough to accommodate the e-commerce it engaged in. Gradually,
the company has become synonym for e-commerce for most of the
products a household may buy – and indeed many businesses also buy
products from Amazon. In addition to books, the company sells many
other categories of product, from entertainment ones (such as CDs and
DVDs), to clothes, furniture and food. Through selling a large range of
products, Amazon has established itself as a strong e-commerce channel for
a large number of people and businesses buying products online. Amazon
has added other services of review and rating to help customers to better
inform their decision. These services rely on user-generated content:
customers give their feedback on products and suppliers, using similar logic
to social media services.
Amazon has become a digital business platform through steps which
transformed the company from an online retail store to an online retail
mall and service provider. For instance, the company changed its business
model to start selling products of small and medium size retailers
(although big retailers can also use the platform). In this way, millions of
small and medium size businesses, which would find it hard to set up from
scratch to sell their products online, have entered the world of internet
sales. Amazon offers the platform for them to sell their products, with the
advantage that they do not need to invest either in the interface (having a
website) or in the security of payment (one of the biggest obstacles for any
company to selling products online). Communication with the customer is
also managed by Amazon and it can also handle the logistics.
In addition, Amazon has become a cloud computing provider. Relying on
its knowledge of how to manage servers, communication and security, it
was a logical step for Amazon to start selling cloud computing services to
third parties. The service supports online operations, meaning the client
does not need to have a server at all to run a particular business. A great
advantage of this model is that the client does not need to invest money
in a fixed infrastructure which may not be necessary for a small business,
at the same time that if or when the infrastructure becomes necessary, it
is immediately available through the cloud. In addition, the system can
be used for back-up databases and big data analytics. And to make the
platform even more attractive, Amazon has the Amazon API Gateway,
which allows developers to create apps to access data and functionalities
that a company is running on Amazon cloud services.
Case 5: WeChat
WeChat is a social media platform used mainly in China. Created by
Tencent, one of the biggest Chinese digital business companies, the app
could be better described as a super-app, which puts together a large set of
functionalities and features, from social media tools (connecting friends)
to instant messaging, purchasing and booking channels and payment
systems, among many other features and functionalities. WeChat has an
environment in which people can organise their digital activities fully. For
instance, if a person wants to book a restaurant or a service, WeChat offers
the features to have access to these services. The same app has the social
networks for organising a dinner, to book the restaurant, to order the food,
to split the bill among friends, to pay the restaurant and to request a cab
to go back home. All in one app (through the so-called mini programmes,
which are apps inside the main app).
201
This centralisation of information in one service provider is a powerful

resource in the hands of Tencent, which knows all that a person is doing
in the day, from their emails and private messages to their photos and
videos, to their conversation with friends, to their consumption of goods
and services, to their savings in the bank, to their health records, to their
entertainment preferences etc. From the user perspective, the convenience
has been a reason for using the app and many would be isolated if they
weren’t using the same platform as their friends (remember also that
Facebook’s and Google’s tools are prohibited in China). However, the
users also become vulnerable to the marketing strategies of a platform
which has all sorts of information about them. Other companies and app
developers have interests in joining the WeChat platform (creating mini
programmes – sub-apps – inside WeChat), as this is the place people are all
the time. Creating an independent app could mean that users would not
adopt the app or would not use the app often for lack of convenience.
Activity 17.4 A new digital business platform for EV

There is, as of today, no large digital business platform to support electric vehicles and
their owners and service suppliers. There are many apps available e.g. to find charging
points, but no global platform as yet.
You want to enter this market and have potentially $3bn available to invest today. Sketch
out in a single page or less your business plan, the architecture you propose and the
services you would provide and to who. Use all you know about infrastructures to help
shape your proposal.

This chapter introduced the concept of digital business platforms,
emphasising how companies in digital industries have become enablers of
third-party business through offering products, services and technologies
which are organised in particular structures. The chapter explained
how digital business platforms also foster innovation, either from
the perspective of supporting new products and services or from the
perspective of new business models. The chapter argued that digital
business platform run on the top of digital infrastructures and ended
giving examples companies which provide such platforms, such as Google,
Apple, Microsoft, Amazon and WeChat.

• explain the relationship between digital business platforms and digital
infrastructures
• describe digital business platforms as digital infrastructures
• identify key digital business platforms
• justify the link between digital business platforms and innovation.
202

1. Define the concept of digital business platforms and how they benefit
from network effects.
2. Explain how digital business platforms are enablers of innovation.
3. Is it justified to see digital business platforms as just another a kind of
digital infrastructure? Are there any clear distinguishing characteristics
you expect to see on digital business platforms?
4. Using the framework of digital business platforms, discuss why Nokia
has lost the competition to Google and Apple in the mobile technology
market (see Gawer and Cusumano, 2013).
203
Notes
204
Chapter 18: Business models and digital infrastructures
Chapter 18: Business models and digital

infrastructures
Chapter outline
This chapter introduces business models and two frameworks to analyse
them. The first presents the key building blocks of a business model, with
a focus on how the parts fit together to form a robust system. The second
also focuses on key elements of business models but proposes a value
perspective aiming to understand better the innovation of business models
themselves. The chapter gives examples of business models, focusing
on companies which provide digital products and services supported by
digital infrastructures. Some of the business models discussed are digital
business platforms. The chapter does not aim to exhaust the discussion on
business models, but to provide a vocabulary and conceptual frameworks
for the student to explore other possibilities, especially thinking on
businesses that are enabled by digital infrastructures.
18.1 Introduction
This chapter presents the concept of business models, illustrated by some
examples. Considering the space limitations of this guide, this chapter
first summarises key ideas presented by Osterwalder, Pigneur and Tucci
(2005), of nine business model building blocks and the implications of
their views for the information system domain, which is our focus in this
subject guide. The content of this part is very close to the Business Model
Canvas, initially proposed by Alexander Osterwalder. You can learn more
online about the Business Model Canvas, which has been well accepted by
practitioners (see further readings below).
Next we present an alternative view on business models, which aims
to emphasise the need of innovating the business models themselves,
as proposed by Rayna and Striukova (2015) (the 360° Business Model
Innovation framework). This second view draws upon the literature on
business models and is quite similarly to the first perspective but it focuses
on the dynamic process of reinventing business models in competitive
environments. The authors argue that in very competitive environments,
such as among companies working with digital goods, it is necessary to
think how to innovate in business models.
It is important to have in mind that, as with any other academic topic,
there are a variety of views on the conceptualisation of business models
and how to apply these models in particular contexts. We present here
two views of this discussion, but you may explore other views through the
literature and future courses.

• Introduce the concept of business models and a vocabulary to interpret
them.
• Describe the nine business model building blocks.
• Describe the 360° Business Model Innovation.
• Provide business model examples related to companies which provide
digital goods.
205

By the end of this chapter and having completed the Essential reading and
• use with confidence the vocabulary to define business models
• describe two business model frameworks and their application in
digital businesses
• explain the application of business models to examples of companies
providing digital goods
• explain the links between business models and digital infrastructures.

Osterwalder, A., Y. Pigneur and C.L. Tucci ‘Clarifying business models: Origins,
present, and future of the concept’, Communications of the Association of
Information Systems 15 2005. Available in the Online Library.

Rayna, T. and L. Striukova ‘360° business model innovation: toward an
integrated view of business model innovation’, Research-Technology
Management 59(3) 2016, pp.21–28.
The link below brings details on the Business Model Canvas, including the
possibility of downloading the canvas and part of the book:
http://www.businessmodelgeneration.com/canvas/bmc
Business Model Canvas explained: https://youtu.be/QoAOzMTLP5s

Osterwalder, A. and Y. Pigneur Business model generation: a handbook for
visionaries, game changers, and challengers. (New Jersey: John Wiley &
Sons, 2010) [ISBN 9780470876411].
Rayna, T. and L. Striukova ‘360° business model innovation: toward an
integrated view of business model innovation’, Research-Technology
Management 59(3) 2016, pp.21–28.

This chapter presents a vocabulary for students to be able to discuss and
design business models. Considering the focus of this subject guide, we have
chosen two academic works focused on analysing companies which provide
digital goods. Although the frameworks are generic, we have selected the
examples which are more relevant for this subject guide. To make it easier,
we give some examples. Towards the end, the chapter connects the idea of
business models to the discussion on digital infrastructures. Although the
frameworks on business models do not apply directly to the study of digital
infrastructures, they are fundamental to understanding the businesses
which are enabled by digital infrastructures. In this fashion, some examples
of business models will be presented, with emphasis to digital business
platforms, linking the example to their related digital infrastructures.

18.2.1 Conceptualising business models
Osterwalder, Pigneur and Tucci (2005, p.4) propose that business models
are ‘the blueprint of how a company does business’, particularly as a
process of translating the organisation strategies into a conceptual model
of how the business should function in order to deliver the strategic
objectives. This business model should show how the company is creating
and commercialising value. This conceptual model would contain objects,
206
concepts and relationships which express the business logic adopted

by the company – how the parts of the business fit together forming a
coherent system. The parts of the business model should be understood
in accordance with the whole. An online action, for instance, is the price
mechanism of a broader business model – even though this element may be
the central part of a business model.
In a more formal way, the authors propose the following definition of
business model (p.10):
Based on this blueprint (a plan), the actual business structure and systems
‘A business model is a conceptual tool that contains a set of
elements and their relationships and allows expressing the
business logic of a specific firm. It is a description of the value
a company offers to one or several segments of customers and
of the architecture of the firm and its network of partners for
creating, marketing and delivering this value and relationship
capital, to generate profitable and sustainable revenue streams.’
are to be created. This is an important aspect of this definition: one needs

to start from the business model plan in order to be able to create the actual
business, for instance, developing the business structure of departments, the
business processes (workflow) and infrastructures and systems (such as the
information system necessary for the business to work).
Naturally, the business model can change through time and then the
organisation structure should be adapted to the new model. The business
model should show how to provide value to customer at the same time that
positive financial benefits are obtained by the supplier of goods and services.
Doing a literature review on the topic, the authors found that scholarly
literature has linked the ideas of business models with digital technologies
from the discussion of transaction cost economics (discussed in Chapter
7 in this subject guide). From the adoption of cheaper information
technology, including accessible broadband and communication services,
it as possible to think new business model, which sometimes would join
different economic agents in the same endeavour. For instance, the iTunes
service (Apple) created a model not only for selling music as a service
but also to sell the iPods and other devices. iTunes also puts together
the interests of Apple and the recording companies, which have been
struggling with the piracy of songs (as discussed in Chapter 7, digital goods
can be reproduced infinitely, opening opportunities for illegal copies being
available free on the internet).
Activity 18.1: Conceptualising business models

There are alternative views on how to define business models. Do research in the library,
consulting books and online resources, to compare different definitions of business
models. Which concepts do you find more useful for understanding how companies gain
competitive advantage from digital infrastructures?
18.2.2 The building blocks of business models

Reviewing the literature, Osterwalder, Pigneur and Tucci (2005) have
concluded that nine building blocks are the most important components of
business models. These blocks are described in the Table 18.1.
The authors consider it important to have a common language to discuss
business models. They argue that the nine blocks described above provides
the common language and conceptual framework for use in the definition
and identification of business models.
207
Pillar Building block Description

PProduct Value proposition Gives an overall view of a company’s bundle of
products and services which create value for a
particular target customer. It should satisfy a
customer’s need through the provision of products
and/or services. For instance, a customer may be
looking for price, speed of service, innovation,
customisation, usability etc. in a particular product
or service.
Customer Target customer Describes the segments of customers a company
interface wants to offer value to. It is also called customer
segments, which may be grouped in accordance
with their characteristics. Any business model
may have one or more customer segments. For
instance, a company may have products for mass
markets and niche markets.
Distribution Describes the various means of the company
channel to get in touch with its customers. This includes
the communication channels and the any
other channel necessary to deliver the value
proposition to the customer, from the moment the
company wants to make the customer aware of a
product, to the purchase, delivery and after sales
communication.
Relationship Explains the kind of links a company establishes
between itself and its different customer
segments. The relationship may range from
personal (face-to-face) to automated (such
as automated emails after a purchase). Some
business models may rely on interaction with
customer in communities and involve the
customer in the process of creation of products
and services.
Infrastructure Value Describes the arrangement of activities and
management configuration resources. It shows the activities and resources
necessary for a business model to work. Activities
may be, for instance, the production of products
and services. Resources may be classified as
physical, intellectual, human and financial
ones. Without such activities and resources,
the company would not be able to create the
products and services and reach the customer.
Core competency Outlines the competencies necessary to execute
the company’s business model. Some authors may
think about this item as capabilities and resource
models.
Partner network Portrays the network of cooperative agreements
with other companies necessary to efficiently offer
and commercialise value. The partnerships are
alliances aiming to maximise the benefits of the
business model, reducing risks and increasing the
access to resources. These alliances may involve
even competitors when all parties can benefit
from it.
208
Financial Cost structure Sums up the monetary consequences of the

aspects means employed in the business model. It
is the sum of the costs incurred to make a
business model possible. They may be fixed and
variable costs. Some companies adopt particular
approaches on the cost structure, for gaining
economies of scale and scope, thus reducing the
cost structure.
Revenue model Describes the way a company makes money
through a variety of revenue flows. It is the money
a company makes from each customer segment.
Each revenue stream may depend on different
price mechanisms, such as actions, subscription
fees, usage fee, brokerage fees and advertising.
Table 18.1: Business model building blocks.
Source: Based on Osterwalder, Pigneur and Tucci (2005, p.10). Additional
information has been added from the same article and Osterwalder and Pigneur
(2010), to clarify details.
In a later book, Osterwalder and Pigneur (2010) use a similar framework
(with adaptations) to analyse the business model of some big companies
offering digital products and services. One of the examples is Google’s
AdWords business model. This service offer companies the opportunity
to publish advertisements on Google’s search pages and affiliated
websites (which publish these adds in exchange for a fee paid by Google
– the AdSense service). The success of the model comes from Google’s
algorithm, which matches the content of the search (and the interest of the
user) to the correct advertisement.
This value proposition depends on having many users searching on
Google, which has not been a problem so far. This flow of searchers
comes because of the efficiency of the algorithm and because Google has
been able to create other services which capture information on users,
improving the quality of their advertising services, such as Gmail, Google
Maps and Chrome (which gives Google access to the browser history too).
In this segment, Google has a revenue model which makes money from
the advertisers, meanwhile it gives free services to web surfers and content
owners (associated with the AdSense service). The search technology,
enabled by the algorithm, is Google’s key resource, which is the foundation
for all three services (Google Search, AdWords and AdSense).
Activity 18.2: Applying the building blocks framework

Using online information, apply the nine business model building blocks to understand
Facebook’s business model. You may face difficulties to find some pieces of information,
but prepare your analysis considering the details you have. Do you know other companies
which have similar business models?
18.2.3 The relevance of business models in the field of information

systems
In this section, we focus on understanding why the concept of the business
model can help us to better understand and design solutions in the field of
information systems. We follow the arguments of Osterwalder, Pigneur and
Tucci (2005), adapted and summarised for the objectives of this subject
guide.
209
A key argument for the authors is that it is necessary to have a good

alignment between the business strategy and the information system
infrastructure which supports the business. It is a fact that a good alignment
between both improves the performance of the business. The business
models favour a better alignment between both the business and the
information technology perspective, as the business models brings more
details on how a particular strategy is going to be delivered. The business
model brings a common understanding on business strategies between the
business and information systems professionals, increasing the chances of
success in creating a better strategic fit between both. In realising how this
alignment could be done, the company would be in a better position to
define their requirements in terms of digital infrastructures.
On a pragmatic level, the business model could help companies to make
better decision on their investments in information technology, as it would
be necessary to think in the long-term strategy of a company, to build a
robust infrastructure. The authors suggest that investments in information
technology should take into consideration each of the nine building blocks,
to explore how digital solutions could improve each of the blocks and the
interaction between the blocks.
In addition, the authors propose that understanding business models and
how they evolve is a form of knowledge management. In describing the
business model explicitly for others to understand, one is documenting a
particular business model that can be communicated to others and changed
in an easier way. As in any situation of knowledge management, the first
step is to convert something that is implicit and tacit into something
explicit. From that explicit knowledge, other solutions can emerge. A
business model is to evolve through time and. Knowing the origin of a
business perspective may help later to change the model when necessary.
18.2.4 Looking for business model innovation

Rayna and Striukova (2016) draw upon similar discussion on business
models, but they are looking from a different perspective: they want to
understand what are the elements that help organisations to innovate in
their business models. They understand that any given organisation has
a business model based on key elements and that these elements may be
changed in accordance with business needs in a context. Both authors are
particularly interested in digital business innovation and they approach the
topic through the lenses of business models. They try to understand why
some companies which are not so advanced technologically can still build
competitive advantages through their business models.
The authors propose a value-based business model framework – 360°
Business Model Innovation – as detailed below, with five key elements and
related sub-components:
Value creation: This is the way a firm can mobilise core internal
competences and resources to create value, using a governance structure to
manage the process for combining competences and resources. A firm also
may use external complementary assets to create value. in this fashion, the
value networks into which a company is inserted is very important, showing
how a company has been successful in building relationships with suppliers
and distributors and involving talents and resources for the process of co-
creation of value. For instance, news companies such as the New York Times
mobilise internal and external resources for creating valuable online news
stories, which are supported by text, photos, images and videos. The sub-
components of the value creation are: core competencies, key resources,
governance, complementary assets and value networks.
210
Value proposition: This is the way a firm defines what is offered in

terms of products and services and for what price. It is not enough to
create value: it is necessary to be able to make this value available to a
particular market by a price that is going to attract the customers. For
instance, a content provider of online news may have a freemium pricing
model, offering part of the content for free, with the objective of attracting
customers interested in subscribing to a service. The sub-components of the
value proposition are: product offering, service offering and pricing model.
Value delivery: This is the way a firm is delivered to the target market
segments through chosen distribution channels. The same company
may have different channels for different customers and/or products
and services. For instance, a book may be delivered by mail or may be
delivered in electronic format to a computer or tablet or electronic reader.
The sub-components of the value delivery are: distribution channels and
target market segments.
Value capture: This is the way a firm can capture the created value. This
should include on the one hand the revenue model and on the other hand
the cost structure of the firm. The difference between both is the profit,
which is allocated across the value chain. In digital economies, the value
allocation is very important. For instance, App Store sells apps for a fee,
retaining part of this fee for Apple (e.g. 30%) and giving the other part
to the app developer (e.g. 70%). Apple uses similar value allocation in
the iTunes service. The sub-components of the value capture are: revenue
model, cost structure and profit allocation.
Value communication: This is the way a firm communicates with
customers and partners about the created value. This aspect covers both
the communication channels (media) and the ethos and story which are
communicated to. For instance, social media channels have helped many
companies to communicate in innovative ways with their customers,
bringing emotions and interactions to the fore. The ethos and story are
also important, because they influence the way others see the company.
For instance, Apple’s Think Different campaign affects the way people
perceive the value created and delivered by the company. The sub-
components of the value communication are: communication channels
and ethos and story.
The authors explain that business model innovation happens when any
of these elements are changed, in such a way that a new combination of
the elements is created. However, they say routine changes in product
offering and delivery channels may not qualify as innovation, making the
ground less clear than one would like. The authors are thus looking for
changes which have a more relevant impact in the market. In other words,
to understand the business model innovation, it is necessary to analyse
the changes inside the firm (focusing on the value elements described
previously) and the result of these changes on the market. Thus, the
authors propose to evaluate the business model innovation considering
both aspects: what happens inside the firm and the impact of the change
in the market. Some changes are modes (sustaining), when the same
customers are kept and others are more radical (disruptive), when new
markets and new industries are reached because of the changes.
Netflix
Let’s consider an example of a disruptive business model: Netflix. The
company was created in 1997, renting DVDs by mail service. Customers paid
per rental. In 1999, the company changed its subscription model, allowing
people to pay a fee for getting the right to see as many DVDs as they wanted
211
in a month. By this stage, Netflix was offering an innovative service available

through an innovative pricing model, changing its value proposition.
A second layer of services, added to the rental one, was made available:
the recommendation of other DVDs based on previous rentals, using a
powerful data mining system, comparing the tastes of the customer with
the rentals of other similar customers (innovation in the service offering).
To improve the quality of its algorithm, the company used an open
innovation model, offering a prize to the innovators who would be able to
improve the quality of the recommendation service (value creation).
Netflix has evolved to become a creator and distributor of original content,
again using the algorithms to produce scripts which would have higher
chances of being more successful (innovation in the product offering).
The investment in the recommendation system has allowed Netflix to
further develop their core competencies, for not only understanding the
consumption patterns but also to understand what the audience would like
to buy, increasing their share in the value chain for changing their position
from being only distributors of content to being also creators of content.
In addition to these core innovations, Netlfix has made other relevant
changes:
• The company has expanded its channels, distributing content
through computers, mobile devices, game consoles and TV sets
(complementary assets).
• It has allowed customers to rate films and series, adding value to all
customers who would like to know the opinion of the audience about
the products (value networks).
• It has diversified the revenue model through selling its own original
content to TV channels.
Through building core competencies and cultivating key resources and
value networks, Netflix has protected its competitive advantage so far in
relation to other big companies which operate in the similar market for
providing digital entertainment, such as Apple, Amazon and Disney Plus.
It is also competitive in relation to traditional TV channels, which have not
been so efficient as to migrate to mobile devices and to use data analytics
to their benefit.
Activity 18.3: Business model innovation

Drawing upon the idea that in competitive environments companies also need to
innovate in terms of business models, analyse how Apple has innovated its business
model from the moment it introduced the iPod (2001), to when it opened iTunes Music
Store (2003) and later when it created the iPhone and the App Store (2008). Think
about the difference in terms of products and service offering, revenue model and value
networks of the different stages.
18.2.5 Examples of business models enabled by digital

infrastructures
The frameworks on business models are not straightforwardly linked to
the idea of digital infrastructures. This is because digital infrastructures
have many layers, with a large variety of companies and organisations
behind each of the layers. Each economic actor has its own business
model operating as part of a large digital infrastructure. The frameworks
on business models apply, however, to businesses that are part of digital
infrastructures or are digital business platforms and to businesses which
are enabled by digital infrastructures and digital business platforms.
212
In this subject guide, we focus on five key digital infrastructures: the

internet, the WWW, cloud computing, mobile technology and social
media. In the paragraphs below, we summarise a few business models of
companies that offer digital products and services, linking their activities
to digital infrastructures. The idea is not to give an exhaustive account of
each business model, but to link the key idea of the business model with
Case 1: eBay and the customer-to-customer marketplace

eBay has innovated in building a customer-to-customer marketplace
(service offering). In order to make this possible, eBay has built a
value network, in which sellers offer their products through the digital
marketplace. The company has targeted a very specific market segment:
the sellers and buyers of used products. Before electronic markets,
this segment was marginal in relation to retail companies. eBay only
distributes the service of the marketplace and is not responsible for the
logistic of stocking and delivering products. The revenue model is based
on commission per sale. The communication channels are all through the
internet, using a proprietary app for use in mobile devices. The company
has complementary assets, such as the PayPal payment system, which
leverages the security of the exchanges for both sellers and buyers.
Case 2: Airbnb and the sharing economy

Airbnb offers an e-commerce service to connect people who want to
rent their home or a room in their home (hosts) to other people (guests)
on short-term contracts. The idea is to offer to guests operating on a
limited budget an alternative to hotel. It is also a good alternative for
guests who need to rent a place for short periods – even if a few months
– in markets where landlords impose long contracts of at least 6 or 12
months. Airbnb is a marketplace for renting property or rooms (service
offering). The company has a revenue model based on commission. The
service is distributed through the internet, having also a proprietary app
for mobile devices. The service is known as an example of the so-called
sharing economy, in which people share with other assets that they do
not need at a particular moment. The target market segments are two:
those who have something to share (in the case of Airbnb, hosts who have
properties or rooms in properties) and those who want to rent somewhere
to stay instead of going to a hotel (guests). The service also includes a
rating system, thus both hosts and guests may rate each other, generating
valuable information for those on the platform (value network), increasing
the value of the interface for all users and for the good hosts and guests.
The rationale of the sharing economy is that individuals or companies
have assets which are not fully utilised. These assets, if shared with
other individuals and companies, can bring to the asset owner an
income, meanwhile that for the users renting the asset, the access to the
benefit can cost less than buying the asset itself. Airbnb starts with this
proposition, of offering a place to be at a fraction of the price of a hotel.
The same logic is used by Uber. In connecting drivers (individuals who
have a car and time to drive the car for another person) and those who
need a ride (cheaper than the regular cab services), the Uber platform
facilitates a better use of available resources, reducing the costs for the
final user and providing income to the driver. There are other similar
services for sharing assets (bicycles, clothes, homes, tools etc.). The
sharing economy reveals that users/buyers are interested in having access
to the benefits of having an asset, not in having the ownership of the asset
213
itself. This rationale is also coherent with an eco-friendlier approach for

modern living, with individuals and families being more prone to procure
access to assets when they need it only. For instance, what is the meaning
of each house in a neighbourhood to have their own lawn mower?
Case 3: Skype and the freemium model

Skype is a Voice over IP service (VoIP), running on the internet (service
offering) from any device, including mobile devices. In order to get a
significant market share of the VoIP, challenging the traditional telephone
companies, Skype entered the market with a freemium model (revenue
model). The company offers free calls for those who are online and can use
the internet to call each other or to do a conference call with more people.
In addition, they offer a service for the same users for calling landline
and mobile phones using Skype. For this service, users need to pay, either
through a model of pay-as-you-go or through a subscription. The fee is still
very competitive in relation to traditional telephone companies. Skype uses
the internet as distribution channel, but it also needs to connect to the other
communication channels when calling landlines and mobiles. The company
has a value network with partners, for instance, to accept online payment
(including payments from PayPal). The information on users and contacts is
supported by a distributed model similar to peer-to-peer (cloud service).
Case 4: Facebook user-generated content and the third-party interest

Facebook is the most used and known social media in most countries,
relying mainly on user-generated content (value network) for delivering
value to its users. The company offers a service of social networking,
allowing people to connect to each other (service offering). The user-
generated content is the product delivered by Facebook (posts, photos,
videos, memes etc.) (product offering). For the users, the service is free.
The revenues come from the advertisers, who want to reach Facebook
users (revenue model). This is a third-party interest model: Facebook
intermediates the access to users, in exchange for a fee, paid by those
who want users to see ads. In order to match properly users and ads, the
company has developed a powerful algorithm (core competence). Facebook
has complementary assets, such as the social games run in its platform,
which also attracts a particular segment of users and Messenger, which
aims to be a substitute for other email services (to a certain degree only).
Facebook also bought WhatsApp, which uses the logic of social networks for
connecting people, but with a simplified interface, which suits many people.
Case 5: Wikipedia and the gift economy

Wikipedia is an online encyclopaedia accessible free from any part of
the world (product offering). Most of the articles in the Wikipedia are in
English, but increasingly versions in other languages are added. The service
is distributed by the internet on any device (distribution channels). The
audience is distributed around the world and could be any person who
wants to know something from a reliable source (target market segments).
Naturally, Wikipedia could not exist without a strong value network: the
contributors who add content free and who keep the content updated
and free from vandalism. They are an anonymous army of knowledgeable
people who dedicate their free time to spreading knowledge (ethos and
story). In spite of having voluntary contributions for the creation of
content, Wikipedia needs money to run the basic managerial structure and
the servers which host its content. This revenue comes from donors all
around the world (revenue model), as Wikipedia does not sell ads.
214
Case 6: Fintech challengers and partners in open innovation

The finance, banking and insurance industries have faced many challenges
from Fintech companies in the last decade. The idea here is to have digital
technologies substituting or changing the way these sectors operate. In
some cases, the challengers have been successful in taking businesses
from the incumbents. For instance, companies specialised in transferring
money between countries have been able to attract customers to use
their online platforms (distribution channels), with the benefit of being
much cheaper and easier to use than banks (value, product proposition).
An example is Wise, previously known as TransferWise. It is extremely
easy to open an account in one of these platforms (just be sure you are
using a safe platform, not a fraud) and to send money to other countries
in other currencies at a fraction of the cost of a regular bank. These apps
demonstrate in practice how much banks have profited from clients who
were forced to use their services. After using these apps, one may feel
outraged by the high price asked by banks to do similar operations.
Other times, the incumbents have embraced technology quickly, to
avoid the challenge, changing their business models. For instance, some
insurance companies are now contracting services from third-party
providers that use AI to calculate the damage to cars after an accident.
This is the case of Vesuvio Labs, a small start-up created in London, which
provides services to big insurance companies. In this example, incumbent
insurance companies, realising they could not compete in developing their
own AI solutions, have decided to partner with Fintech companies such
as Vesuvio Labs (value networks). The emergence of Fintech applications
and companies is forcing change in the finance industry, with new business
models emerging. For instance, incumbent banks are more prone to
adopt an open innovation business model, to facilitate the acquisition
of digital solutions from third parties, instead of developing solutions in
house. In the past, incumbents would be more prone to develop solutions
themselves (core competences, key resources). However, as Fintech
technology is advancing quickly, an open innovation approach may
allow the incumbent to innovate with less investment in research and
development. Those who adopt an open innovation business model are
perhaps more likely to survive the market competition.
Fintech is changing the infrastructure of banking around the world,
as financial institutions are reinventing their business models. An
interesting example is the platform WeTrade, commissioned by a group
of major European banks and developed by IBM. The objective of the
platform is to foster international trade among companies (buyers and
sellers), connecting them through their respective banks. The platform
uses blockchain technology, to make transactions safe and immutable.
Companies benefit from having in the same platform access to secure
transactions with any other company in the system (increased level of
trust, provided by the platform, as only legitimate businesses are allowed
into the system), reducing risks and costs and improving the management
of information. For the banks, this is an opportunity to offer added value
to business, also increasing their loyalty and profitability. Individual
businesses would not be able to create such a complex platform, but a
consortium of banks have the economic power to contract IBM for the
development of the platform.
Fintech based on AI (e.g. machine learning) is also rapidly changing
many aspects of banking. For instance, AI is used to identify frauds in
financial operations, through observing individual patterns of behaviour
and comparing with databases of previous behaviour of the individual
215
and of all customers. AI is also used for voice recognition systems (such
as Natural Language Processing), which allow customers to communicate
with the banks with voice-command. Although this technology is still
precarious (particularly when customers have accents which are not
recognised by the AI), it is improving very quickly. Sentiment analysis is
also used by banks and financial institutions to improve the quality of their
customer services. Sentient Machines, for instance, is a company providing
fintech solutions based on AI, for recognising the feelings of the customers
when contacting the bank. The system can, for instance, identify when a
call is urgent, to give the customer the right level of priority and the access
to the right professional.
Activity 18.4: Linking business models with digital infrastructures

Drawing upon the six examples provided in this section, think about two other companies
which provide digital services and/or products online. Identify key value elements of their
business models. Link the business model to digital infrastructures.

This chapter introduced the concept of business models, presenting two
frameworks. The first framework – the nine business model building
blocks – is presented as a good way to understand the nature of business.
It has been argued that using a clear business model brings benefits for the
alignment of business and information technology. The second framework
– 360° Business Model Innovation – is presented as a conceptual
framework to understand the innovation of business models themselves.
Some examples have been constructed to show the links between business
models and digital infrastructures.

• use with confidence the vocabulary to define business models
• describe two business model frameworks and their application in
digital businesses
• explain the application of business models to examples of companies
providing digital goods
• explain the links between business models and digital infrastructures.

1. Describe each of the nine business model building blocks proposed by
Osterwalder, Pigneur and Tucci (2005).
2. Explain why the discussion of business model innovation is relevant in
contemporary economies.
3. Explain as clearly and briefly as you can why the topic of business
models is related to the topic of digital infrastructures. Give examples
supporting your argument.
216
Chapter 19: Smart cities
Chapter outline
This chapter explores the concept of smart cities to show how digital
infrastructures can be combined with physical infrastructures and
embedded in communities, in order to improve the management of cities
around the world. The chapter discusses some challenges for smart city
solutions and the cases of Singapore and Amsterdam, two examples of
smart cities around the world.
19.1 Introduction
The term smart city has been used to define urban developments which
integrate information and communication technologies to better manage
cities and improve the quality of life of people in cities. The idea is
to have digital technologies integrated with the life of the city, from
government departments and transportation systems to schools, hospitals,
entertainment and leisure, energy and water supply networks and waste
management, among other areas. Being more efficient and capable of
controlling outputs, smart cities aim to save public money but more
importantly to and deliver better services to citizens. The concept of smart
cities is very broad and used in different contexts to represent a large
range of possibilities in terms of using digital infrastructures to manage
urban spaces and serve citizens.

• Introduce the concept of smart cities.
• Explore the integration between digital infrastructures and city
services in smart cities projects.
• Describe examples of smart cities solutions.
• Discuss challenges of smart cities solutions.

• use the concept of smart cities in proper contexts
• describe key digital infrastructures which support smart cities projects
• explain key applications of smart cities solutions, understanding
benefits and costs
• analyse the challenges related to smart cities projects.

Smart city: https://en.wikipedia.org/wiki/Smart_city
Smart Nation Singapore: https://www.smartnation.sg/
Amsterdam Smart City: https://amsterdamsmartcity.com/

Fitzgerald, M. ‘Data-driven city management: a close look at Amsterdam’s
smart city initiative’, MIT Sloan Management Review (updated May 2016)
https://sloanreview.mit.edu/case-study/data-driven-city-management/
217
Saunders, T. and P. Baeck ‘Rethinking smart cities from the ground up’
(18th June 2015). NESTA Publications. https://www.nesta.org.uk/report/
rethinking-smart-cities-from-the-ground-up/

The chapter conceptualise smart cities, emphasising the adoption and
adaptation of digital infrastructures in the scale of urban developments.
The chapter presents key examples of technologies and key challenges for
smart city projects. The chapter describes key features of Singapore and
Amsterdam smart city projects, which are two examples of how digital
infrastructures may improve the management of urban spaces.

19.2.1 Conceptualising smart cities
Smart city is a term used to express the integration of two major aspects
– digital technologies and urban developments – in order to provide
better and more efficient services and quality of life to citizens. This is
done through investments in technologies which enable the more efficient
management of data and information collection and flows. The integration
may involve also the idea of communities and how digital technologies can
foster better communication and decision making driven by communities,
for instance, when citizens are called to consult on decisions directly,
instead of through elected representatives. Considering that increasingly
people live in urban developments, it is urgent to find means to make
these spaces more sustainable and efficient.
Different cities integrate information and communication technologies in
different ways. For instance, some may invest more on ICTs for controlling
traffic and transport. Others put emphasises on the management of energy,
water supply and waste management, aiming for more sustainable urban
developments. Others adopt information and communications technologies
to better manage common services such as schools, hospitals and social
care. The idea of sustainability, environmental improvement and a greener
footprint are ambitions are cited many times when discussing smart cities,
although in practice there are many initiatives which are not strongly
connected with such efforts. Thus, in this subject guide we do not limit the
concept of smart cities to those who emphasise the environmental benefits
of getting digital technology for supporting the management of cities.
The concept of smart cities involves the integration of at least three layers:
the physical infrastructure, such as roads, hospitals and water supply
pipes, the digital infrastructure (which includes physical elements such
as the broadband backbones and the internet of Things, IoT and software
solutions) and the logical and institutional infrastructures (such as the
management of information flow, the generation of big data and the use of
data analytics). More sophisticated projects may include human and social
capital layers, understanding that smart cities improve the quality of living
in cities and that digital technologies may enable better communication
and collaboration between citizens, thus affecting the way communities
live in cities. In some cases, the digital infrastructures and related
applications aim to improve the intelligence of the city, either because
better decisions are made by the collectives, or because layers of artificial
intelligence can be added to better use the data available. Some smart
cities projects aim even to foster innovation clusters. An additional layer
could be the creation of resilience, for cities to adapt to new challenges
such as those of climate change, natural disasters and security threats.
218
As you would expect, smart cities evolve over time on the basis of
developing underlying infrastructure and allowing innovations in how it
is used. The city needs to be connected by a digital infrastructure such
as the internet (fast broadband) and have wireless access everywhere, to
collect the necessary data. Electronic sensors and RFID (Radio Frequency
Identification) can be connected to things which can generate data, thus
building an IoT infrastructure. Data analytics can propose which pieces of
information are relevant and how to analyse these by mathematical and
statistical models and AI analytical tools (using machine learning).
Activity 19.1: Conceptualising smart cities

Think about one difficult and enduring problem in a city you know. Develop a quick
overview of a smart city project which might help the problem by integrating digital
infrastructures in the urban space. Consider any problems in the identification, capture,
aggregation and analysis of relevant of data?
19.2.2 Types of smart-cities projects dependent on digital

infrastructures
This section describes some digital infrastructures related to smart cities
projects. In this section, we explore the variety of possibilities. Later, we
discuss examples of cities which have integrated some of these projects
in initiatives to build smart cities. This list is not exhaustive, as there are
many other possible applications of digital infrastructures in smart cities
projects.
Traffic management systems

These systems aim to measure how the traffic evolve through the day,
to give guidance and redirect traffic to routes which are better options
for vehicles (reducing the waste of time and fuel in the transportation
system). In an emergency, the system can change the necessary traffic
lights to green for allowing emergency vehicles to reach their targets
faster and in a safer manner (avoiding accidents with between emergency
services and other vehicles). This sort of traffic management system
depends on digital technologies such as sensors, GPS, RFID, wireless
networks and internet infrastructure. The information can be collected
from fixed points in the roads, which would measure the flow of cars,
or from the vehicles themselves, considering their movement and GPS
location. Through mathematical modelling (using previous and current
data), it is possible to detect earlier when a congestion will happen, thus
allowing the system to adjust itself as earlier as possible to get a better
distribution of vehicles through alternative routes. The system may also
be integrated with drivers’ mobile phones, sending real time information
about which routes drivers should take or avoid and about the best time
to move from one place to another (including alternative transport means,
from private cars to public transportation).
Parking spaces
Related to the problem of traffic, there is the problem of how to find a
parking space in high populated cities, especially in central areas, where
parking is more controlled and difficult. Information on parking spaces
can be measured by sensors. Again, the information would be aggregated
centrally and informed back to those who are interested in finding a
parking space, considering their location (informed by GPS systems). The
matching of parking space and vehicle looking for a space can be done
219
through public electronic displays or mobile phones, for instance, if a driver

has a particular app, or through SMS systems. Private and public parking
spaces could be included in such a system. For instance, private parking could
become publicly available when the owner of the space is not using it. Even
better, the owner of such a private parking space could make some money for
allowing people to park there. This management of parking spaces could also
perhaps have a positive impact on pollution, as people would spend less time
looking around for a space. On the other hand, if we take a longer-term view,
our cities probably need less traffic and cars. So perhaps the ambition should
be for smart public transport, not more parking.
Management of pollution and noise

Big cities face unavoidable levels of higher air pollution and noise. Sensors
can be distributed around the city to measure the levels in each critical
point of the urban space. With this real-time data collection, it is possible to
use the wireless and broadband network connections to send information to
a central database, which aggregate the data and informs decision makers
on the best alternatives. For instance, a city may prohibit private cars in the
streets in a critical situation in which the level of pollution is not acceptable.
Schools may be closed for periods in which it is not safe for children to be
on the road because of the air pollution. Emergency services and hospital
may be prepared to take more patients in critical conditions because of poor
air quality. The sort of measure depends on the nature of the problem. The
focus here is on understanding that information is collected and aggregated
quickly to inform decision makers.
Government services
This item includes all forms of services that can be improved and be more
efficient when delivered through digital infrastructures. For instance, city
governments may offer online information about all services provided,
booking services, such as the collection of waste or the cleaning of a
public area and online access to the electoral register. They may also offer
payment interfaces, for instance, allowing payment of council/city taxes
online, without going to a bank. The focus here is on the management
of governmental services which are provided by other means and gain
efficiency and quality through the redesign of solutions supported by digital
infrastructures, platforms and applications, such as the internet, the IoT
and big data. E-services also target reducing the amount of paper used,
thus having an element of green IT solutions. In order to make this access
easier to citizens, local and national governments may issue smart cards, i.e.
electronic identification cards which allow citizens to have access to many
e-government services using the same online identification. This enables
citizens to avoid having to register for different services and manage different
accounts when using services provided by the same government office.
Social and healthcare services

Digital infrastructures may support a large variety of social and healthcare
services. In hospitals, for instance, digital technology may improve the
quality of information collection and flow. This can be integrated with the
general practitioners and labs, allowing the data from a particular patience
to be immediately available in the whole health system, thus speeding up
decision making and reducing mistakes, such as giving the wrong medicine
to patients. The system may also gain efficiency through integrating the
ambulance services, not only to rescue patients quicker but also to find
the best hospital for them while the ambulance is on the road. Based on
the constant flow of data, algorithms may analyse real-time information
220
to adjust the flow of services, for instance, quickly redirecting ambulance

services to attend a large critical event with many victims, reducing services
for less critical situations. Smart city solutions may also connect a network
of doctors, nurses and health services 24/7, allowing the system to find the
closest person to act in critical events, matching the information they have
about the patient and the localisation of the health professional in real time
(through GPS).
Large scale digital infrastructures can supply information on the health of
populations, for instance, speeding up the process of identifying epidemics
and collective health issues. Online interfaces can give support to patients,
families and health professionals, including interfaces which rely on AI
and machine learning to improve the results of consultations. From the
perspective of social care, aging populations need more support which
cannot be given by humans only. Wearable devices, for instance, may extend
the independence of older citizens, through monitoring their movements.
When the device identifies abrupt falls or lack of movement, health services
can be informed to take immediate action. Mobile technology may be added
to the routine of social care and healthcare professionals, who can collect
and record data quicker and with fewer mistakes (common when data is
transferred from a paper system to an electronic system). Citizens can have
better access to social care and healthcare systems, when they use their
mobile interfaces to get information or request help.
Management of water, energy and waste

As in other examples, sensors measure a particular performance metric.
For instance, sensors could capture information that the consumption of
electricity in the city has reached critical levels, recommending planned
power cuts of short duration or increasing the flow of electricity to the
locality. Without real-time information there are higher chances of blackouts
happening. Similarly, controlling the flow of water in crucial points in the
city may not only inform whether the levels of water are sufficient for a
particular demand, but also to inform whether there is an unusual flow of
water in a neighbourhood, which could indicate a burst water pipe. High
levels of water consumption during a draught may be an indication that
the government needs to warn households and companies on the risks
of overusing water. In addition, sensors may help to warn of flooding in
critical areas, anticipating the evacuation of people when necessary and
problems with the quality of the water, such as contamination by chemicals
or microorganisms. Sensors in parks and green areas may inform when
the plants need water, thus starting automatic irrigation which deploys the
correct amount of water.
In terms of waste management, sensors inside bins inform collectors of
those containers which are full and need immediate cleaning to avoid
overflowing. In this way, the collectors maximise the benefit for the city,
always reaching the more critical areas faster than previously. Integrating
this system of waste management with the traffic system helps a service
provider to not only cope with the waste but also to be more efficient in
the use of vehicles for collecting waste. The planning in this case may be
done by algorithms which calculate the best rota to collect waste depending
on the combined needs of the city in terms of waste and traffic situation
together.
Smart sensors may reduce street lighting when there are no people or
vehicles on a road, or turn off heating and lighting in empty buildings. In
houses, smart meters can help families to better manage their energy use,
identifying the drivers of energy bills and ways of reducing costs, such as
221
programming activities for periods when energy is cheaper. Legislation

may enforce the adoption of similar solutions in cities concerned in find
more sustainable solutions, as well as saving money.
Electric vehicles infrastructure

In face of the pressures to reduce carbon emissions, there is great
expectation that electric vehicles are going to replace those using carbon-
based fuels (e.g. diesel, gasoline). These vehicles, in addition to being
powered by electricity stored in batteries, have the advantage of having
a high degree of embedded digital technologies, which control their
performance (passing information on requirements for maintenance,
for instance) and provide self-driving features (in different degrees of
development). Although there is still need of further development of the
digital features for producing fully autonomous vehicles and although
it is necessary to have a legal framework at a national level to allow
such autonomous vehicles to operate without human supervision, the
expectation it that eventually the norm will be to have fully autonomous
vehicles, run on electric energy.
To allow this development to come, countries and cities must invest
massively in the infrastructure for the electric vehicles, for public and
private transport. A critical bottleneck for the adoption of electric vehicles
is the lack of infrastructure for charging them. Currently, people charge
their vehicles at home but the availability of public charging stations is
low. Cities face this chicken-or-egg dilemma: there is little investment
in public infrastructure for charging stations as we still do not have
many electric vehicles, but people do not buy electric vehicles as they
think about the inconvenience of not having enough charging stations
when they need them. One of the sides needs to face the risks of the
investments. Considering the product lifecycle of individual cars and of
the charging infrastructures, it is more difficult for final consumers to foot
this bill (shorter lifecycles). Thus, cities need to invest in the infrastructure
first, to motivate more people to have electric vehicles. This infrastructure
will also support public transportation projects involving electric vehicles.
Additionally, as electric vehicles bring the most recent digital technology,
connecting them to the internet through wi-fi, they are easily integrated
into other smart city projects, for controlling traffic, parking and pollution,
for instance. The higher the number of smart vehicles, the easier to
implement traffic and pollution controls, as cities may receive direct
information from vehicles (instead of investing more on external sensors
distributed through the city, the IoT). The vehicles themselves have the
sensors to pass information, as far as they are integrated into the digital
platforms. For instance, cars may pass information about traffic to control
rooms. Also, if an area has more electric cars than another, it will have
less vehicle pollution. These cars also are more easily connected with
infrastructure for parking, being able to identify the closer available spots,
or infrastructure for road pricing – charging vehicles for using specific
roads at specific times. There is an opportunity then for smart city projects
to work with car manufacturers and the owners of electric vehicles to
better integrate the cars into the digital infrastructure of the city. There
are apps which coordinate the car sharing, for people to share their cars
with others who do the same trajectory. Smart city projects can better
coordinate these efforts with a more robust integration of these cars into
the digital information infrastructures, fostering the sharing economy, e.g.
using the same stock of public and private assets to generate more value
(more and better transportation, less pollution and lower costs). All these
222
potential benefits from creating a physical and digital infrastructure for the
electric vehicles, integrating them into the data management systems, can
be enabled by smart city projects.
Activity 19.2: Smart city security

Historical and real-time data, associated with data analytics, may help police institutions
to predict where crimes are to occur in a city. Using smart city solutions, police can better
deploy resources in the areas where there is a greater probability of crimes occurring.
Research online how similar solutions have been deployed in two or three cities around
the world. What do you think about these solutions?
Activity 19.3: Smart city crowdsourcing

Social media channels help citizens to express their opinions and concerns. Some smart
city projects propose to use social media to better capture in real time the opinion of
citizens on the management of city problems. The idea behind these projects is the belief
that the wisdom of the crowds can be capture by social media, better informing local
governments on the most important problems and solutions for a community. Research
online examples of cities that have been using social media to engage citizens in the
discussion of community problems and solutions.
19.2.3 Challenges and concerns

There are some concerns on whether smart city solutions should be
adopted. Three main counter arguments are presented below, but of
course there are potentially more in any specific circumstances.
Smart city solutions demand high investments. Local governments need to
invest substantial amounts of resources to make possible the development
and implementation of digital infrastructures to support smart city
projects. Some critics think the cost is bigger than the benefit and that
the money could be used in better ways. They say digital technologies are
evolving very quickly, thus new developments may make recent solutions
obsolete, leading to spiralling costs which are beneficial for the public
finances of the city. On the other hand, those who defend smart city
solutions argue that when done properly, the business case is favourable
to smart city projects, as the money saved is bigger than the investment, in
addition to gaining better-quality services. Part of the investment for smart
cities, such as high-speed broadband and wireless facilities are to benefit
not only the government but also the whole community. In democratic
countries, citizens may question decisions that may not be the most cost-
effective way of solving city problems.
Smart city solutions increase the dependency of governments on
particular vendors. This challenge is not only for smart cities projects,
but for any governmental project involving large-scale investments in
digital infrastructures and technology. The concern here is that once a
government starts using a particular vendor as the main provider of a
solution, it is very difficult to change the provider later. There is thus
path-dependency in relation to the adopted technology, which empowers
the vendors in the long term. Some vendors may offer lower prices to
get key initial contracts with governments, understanding that once the
city depends on a particular platform, it would be very costly to change
providers. Thus, the bulk of costs for a city is never in the initial contracts,
but in the later contracts when there is dependency of a particular solution
provided by a vendor. To avoid this sort of problem, some governments
have adopted a more fragmented approach to the purchase of digital
technology, combining solutions provided by vendors with solutions
223
designed by public service professionals, who may also rely on open-source

software. However, the integration of hardware and software solutions and
the difficulties in managing fragmented systems may be an obstacle for this
sort of strategy, with hidden costs building up.
Smart city solutions increase the risk of surveillance of citizens. Smart cities
are about collecting and centralising data and information, which may
reveal more about the behaviour of groups and citizens. Depending on the
way the projects are implemented, data from individuals may be gathered,
increasing the chances of governments controlling citizens, if the political
institutions of a country would allow such a control. For instance, the use
of digital cameras in cities such as London and New York improves security;
however, all people are recorded. What happens if a government has the
means to know, for instance, when each person goes to work or should go
to work or school? What happens when governments have information
on people’s movements and expected behaviour? Perhaps in a democratic
country, this concern is never going to materialise. However, the digital
technology used to support smart cities can also be used to increase the
surveillance of citizens, against their will.
Smart city solutions increase the risk of cybercrime. The digital
infrastructure supporting smart cities may be hacked by criminals, as
any other digital infrastructure, which would make citizens vulnerable
to having their privacy violated. Organised crime groups could explore
the vulnerabilities of such systems. For instance, criminals could check
information from smart meters to decide which houses are to be burgled
(giving preference for entering houses when occupiers are absent).
Or knowing details on the privacy of individuals, criminals could use
the information to demand ransom payments to not make a piece
of information public. In countries in which violence is high, private
information about individuals may be used to kidnap people. The
government itself could be blackmailed, if organised criminals take control
of digital systems which control traffic or hospitals.
Smart city solutions may exclude social groups from services. When a
government migrates services to digital platforms, those citizens who
are not digitally literate, may be excluded from services or may be in
disadvantage in relation to other citizens when accessing the same
services (for instance, having to take more time in queues to book services
which could be otherwise easily booked online). The fact is that many
governments and companies have adopted the policy of making offline
interactions more difficult, to force people to use digital platforms of all
sorts. If one does not want to use a digital platform (or does not know
how to use it properly), one is excluded from many benefits or is forced
to pay a higher price to have the same benefits of others. From an ethical
perspective, governments must be more inclusive in their smart city
projects, to avoid exclusion of social groups. This includes keeping offline
services with the same level of quality of online ones, to be more inclusive
of all social groups.
19.2.4 Singapore and Amsterdam: Creating smart cities
Singapore
Singapore launched in 2014 its Smart Nation programme, aiming to
become one of the most interesting examples of how digital infrastructures
can support the emergence of a smart city. The government promises
to deliver better quality of life, greater integration of communities and
urban developments which are sustainable across generations. The
224
official guidelines say the adoption of smart city solutions will improve
government services through digital infrastructures. In addition, the
government wants the investments to foster innovation. To support this
aim, it has created the Jurong Innovation District, which brings together
researchers, entrepreneurs and students to develop products and services
coherent with the proposition of smart city projects.
The Smart Nation programme is possible because the government
systematically collects data about the daily activities in the city, through
sensors and cameras. The collected data is aggregated in an online
platform called the Virtual Singapore, which allows the government
to observe in real time how the city is working. Data collection and
processing enable services. For instance, connected citizens are informed
about public transport, traffic and parking spaces in real time, allowing
people to crowded parts of the transport network and choose better
driving routes. Gradually, the digital infrastructure of Virtual Singapore
should evolve to more ambitious goals, such as controlling how infectious
diseases spread through the population and how crowds react in the event
of an explosion in a public space. Drivers could have additional sensors
in their mobiles to measure the movement of their cars in the streets,
informing the government when a road needs maintenance. Sensors
connected to elderly citizens may inform if they lose consciousness or need
help or healthcare.
From a legal perspective, digital infrastructures can support law
enforcement, although sacrificing privacy. For instance, by controlling
whether people are smoking in prohibited areas or littering public spaces.
Under Singapore laws, the government can install sensors and collect data
for reasons of law enforcement or even simply surveillance of citizens.
The Singapore government owns most of the buildings in the city, so it is
very easy to make decisions to add sensors and cameras. These sensors
can measure everything about the building (materials, dimensions), but
they also measure everything about the way people move in the building
and use their own space. For instance, a person could be photographed
throwing cigarette butts in the street, allowing the government to fine the
individual for disrespecting the law.
The Smart Nation programme is evolving gradually, and it is not clear how
data will be used in the future. For some people, this system empowers the
government to increase the surveillance of society. Other people agree that
the government has the right to invigilate citizens, to be sure the law is
respected.
Amsterdam
Amsterdam Smart City programme was launched in 2009. One of the
goals of the programme is to look for a balance between the visitors and
the residents of the city, and between the need of economic growth vis-à-
vis the necessity of fostering long-term sustainability. The programme has
dozens of projects across the city, many of them at the testing/pilot stage.
A key pillar of the programme is the open source data approach. The
government has made an investment to consolidate and clean thousands
of datasets from different departments, to create a meaningful, reliable
dataset. Individuals and companies can access public open source data to
create applications and services. This data is provided in real time, through
digital infrastructures, and developers can use an open API to create
solutions using public data. For instance, developers may use the data
about transportation systems and traffic to guide citizens about the better
225
public transport and road options at any given moment. Data analytics
and data modelling on traffic can be used to reprogramme the traffic lights
during the day according to traffic flow needs. This smart solution can
reduce traffic substantially, through adapting the model in real time in
accordance with the most recent information.
There are projects which rely on the help of citizens to collect data on
air pollution and noise, through sensors attached to mobile phones. On a
large scale, the government can map energy consumption across the city,
enabling it to propose better use of solar energy in the neighbourhoods
that need it most. The energy generated during the day can be stored in
electric car batteries, which can provide energy to households as necessary.
Smart meters support better energy management in residences and offices.
In addition to products and services, Amsterdam invests in the creation
of Living Labs. The idea is to create small clusters of each product and
service, which are tested, before replicating for the whole city and
beyond. For instance, projects testing personalised television systems and
transportation services, or researching the impact of co-working spaces in
the use of energy and generation of pollution. These co-working spaces
depend on efficient digital infrastructures to allow employees to work
remotely and efficiently. The idea of fostering the engagement of citizens is
behind some of the smart city projects in Amsterdam, in order to get more
people participating in knowledge sharing and innovation, in addition
to influencing public policies. For instance, the Amsterdam Institute for
Advanced Metropolitan Solutions sponsored the project Beautiful Noise,
analysing social media data posted by residents and tourists to capture in
real time events such as long queues in museums and formation of crowds.
Activity 19.4: Finding other examples of smart cities

There are many cities around the world using digital infrastructures to transform the
management of urban space and urban life. Research online at least two cities which
have transformed themselves into smart cities. Discuss the advantages and challenges
found in these two settings.

The chapter discussed the concept of smart cities in order to demonstrate
how digital infrastructures may be integrated in large-scale urban
developments, which aim to bring more efficiency and quality to the
management of cities. The chapter presented key examples of technologies
and key challenges for smart city projects. Finally, the chapter described
key features of Singapore and Amsterdam smart city projects.

• use the concept of smart cities in proper contexts
• describe key digital infrastructures which support smart cities projects
• explain key applications of smart cities solutions, understanding
benefits and costs
• analyse the challenges related to smart cities projects.
226

1. Explain how and why digital infrastructures are fundamental for the
development of smart city projects.
2. Discuss the main benefits and challenges when developing smart city
projects.
3. Present three examples of smart city systems which you think are or
could become fundamental for improving the quality of living in your
city.
227
Notes
228
Chapter 20: Reflections on the course: the future of digital infrastructures for business
Chapter 20: Reflections on the course:

the future of digital infrastructures for
business
Chapter outline
This chapter summarises the key arguments of the course and prepares
you to study and research further related topics. The first objective is to
help you understand how the whole course comes together and to allow
a better understanding of the big picture of digital infrastructures for
business. A second objective is to motivate you to keep up you research
on related topics, as innovation in digital infrastructures is opening new
business frontiers every day.
20.1 Introduction
Our goal in this course is to provide a broad understanding of digital
infrastructures for business. As this is an introductory course, we have
been careful to add gradually more details on complex topics, which will
be developed further in later courses. We have balanced the technical
and the businesses aspects, giving you enough information to form a
view of the interplay between the two aspects of digital technologies.
Even for business students, it is necessary to have a degree of technical
knowledge in order to better understand how digital infrastructures evolve
and enable new opportunities. This knowledge also will help you to have
better conversations with technical colleagues in business environments,
increasing your chances of having more productive and successful
collaborations.
We have balanced the content of this subject guide to cover facts and
theories, again taking into consideration that this is an introductory course
so a selection of relevant theories and models have been introduced and
then applied to topics. It is important that you understand how such
theories can give more depth to your analysis of digital infrastructures and
better prepare you for the job market in the future. As we hope you have
learnt from this course, innovation in digital technologies advances quickly,
with serious implications for organisations, business sectors, individuals
and societies. We can get lost if we focus only on the technical aspects of
innovation and uncritically accept the latest gadgets or business models.
In this rapidly changing domain theory (or rather theories) can help you
understand more and make better judgements as innovations arise.
In this chapter we provide a summary of key ideas and theories discussed
in this guide and then introduce some emerging topics for your further
research and reflection. These topics are chosen because they are
innovations that are being created and first put to use now and some of
them they may become significant in your professional life hereafter. It’s
best to be prepared for the challenges.

• Present a summary of the subject guide, emphasising key technical,
business and theoretical aspects related to digital infrastructures for
business.
229
• Point out key areas in which digital infrastructures are evolving.

• Link the current development of digital infrastructures with the
challenges faced by organisations, individuals and societies in terms of
education and career plans.
• Motivate you to perform independent research on the topic, engaging
in a transformative process of learning how to learn.

• explain how the parts (chapters) are combined into the whole course
• describe current innovations that draw on and exploit digital
infrastructures
• explain how such innovations challenge organisations, individuals and
societies
• outline individual and institutional strategies for continuously learning
to adapt to changes in a context of constant innovation in digital
infrastructures.

Big data: https://en.wikipedia.org/wiki/Big_data
Algorithm: https://en.wikipedia.org/wiki/Algorithm
Artificial intelligence: https://en.wikipedia.org/wiki/Artificial_intelligence
Machine learning: https://en.wikipedia.org/wiki/Machine_learning
Internet of Things: https://en.wikipedia.org/wiki/internet_of_things
Watch the videos below:
Andrew McAfee at TEDxBoston ‘Race against the machine’: www.youtube.
com/watch?v=QfMGyCk3XTw
Andrew McAfee ‘Racing with – not against – the machine’: www.youtube.
com/watch?v=IqBK6E7mAuc
Andrew Mcafee: ‘The second machine age’: www.youtube.com/
watch?v=KoC63M33lvw
2014 ‘Keynote closing in on the second machine age’: www.youtube.com/
watch?v=AX4BbRgMpwU

Brynjolfsson, E. and A. McAfee Race against the machine. (Lexington,
Massachusetts: Digital Frontier Press, 2012) [ISBN 9780984725113].
Brynjolfsson, E. and A. McAfee The second machine age. (New York: W.W.
Norton & Company Inc., 2014) [ISBN 9780393239355].

This chapter summarises the key discussions of this subject guide. It
reminds you about key models, conceptual frameworks and theories, such
as the layered model, ideas on standards and network externalities, and
macro perspectives on the development of digital infrastructures based
on economics and institutional theory. The chapter also emphasises the
connections between the digital infrastructures and the way they are
related to business innovation (digital platforms, business models, and
digital products and services). In addition, this chapter finishes with two
230
sections, one to explore further ideas on how digital infrastructure may

evolve considering recent technological developments, and the second
to discuss the impacts of the pervasiveness of digital infrastructures in
contemporary economies and societies.

20.2.1From digital infrastructures to digital business innovation:
what we know so far
This subject guide has presented a selection of topics and theories to help
explain how digital infrastructures evolve and how they are connected
to digital business innovation. We have discussed technical and business
aspects of digital infrastructures and we have also selected key models,
frameworks and theories which can help us to better understand the
emergence and relevance of digital infrastructures.
We adopt the layered model as our most fundamental conceptualisation
of digital infrastructure and in particular the hourglass architecture.
We have chosen this model as a good representation of the way digital
infrastructures evolve and how they enable innovations. As they are not
designed from scratch, digital infrastructures evolve through time, building
upon previous layers of technology – the installed base. Each layer brings
a diversity of elements which are adapted as necessary over time to
support the many data, information and business processes (applications)
we observe as users of such systems. The hourglass architecture helps us
understand that we do not need to have a big design plan to achieve a
huge digital infrastructure. The best examples are the internet and the
WWW. The history of these infrastructures show how something initially
used by a few researchers has become a massive digital infrastructure for
the whole world. It is hard to imagine our societies and economies without
the internet and WWW.
If we do not design digital infrastructures from scratch, how is it possible
that the parts come together? We have discussed that the most important
mechanism is to allow the layers of digital infrastructures to work together
in the process of agreeing standards. For instance, the internet works as
one network – in spite of being a myriad of individual networks coming
together as one – through the use of communication standards such as the
TCP/IP. Each network has different hardware and software behind, but it
is connected to the whole through the TCP/IP. This allows the individual
networks to evolve without losing the connection to the whole. Through
standards, different pieces of technology interact and communicate with
each other, forming a whole.
The use of a layered architecture and standards shows that the
development of a digital infrastructure has a path dependency in relation
to the installed base. All technical and institutional decisions made in the
past decades will influence the way digital infrastructures evolve today
and the way they are able to evolve in the future. There are developments
that may just not be possible because of the installed base. On the other
hand, the layered architecture and the use of standards gives us a level of
flexibility to develop innovations, improvisations and workarounds. So the
installed base and path dependency are offset by digital generativity and
the potential for innovation.
Innovation indeed is a crucial concept in this course. This subject guide
has emphasised two aspects of innovation. The first is the innovation of
digital infrastructures themselves. For instance, when faster networks
231
allow the streaming of videos on the internet on a massive scale, or when

computer programming allows the creation of cognitive computing (AI).
These innovations can be added as layers of digital infrastructure, sometime
becoming digital business platforms which enable new business activities.
The second aspect is innovation in business models and the creation of
new digital products and services within new market structures. We have
emphasised across this subject guide that this second aspect is a core
responsibility of business professionals and we have presented theories to
support our analysis. The discussion on digital innovation tends to focus
on the technical aspects, but we are in a context in which we equally need
to think of innovation in terms of new users and new uses – new digital
business models. In the end, the most important innovations come from the
way people and organisations appropriate digital technologies into their
lives.
Lastly, still from a theoretical angle, we have introduced some economic
and institutional perspectives which can help us to better understand why
and how digital infrastructures have evolved. These theories focus on the
logic of growth, for instance positive network externalities and reducing
transaction costs, or in institutional influences on society. These topics will
be discussed in more depth in future courses in your programme. The basic
message from this course is that there are economic theories to explain
how and why digital infrastructures have become so pervasive in business.
We also discussed a more strategic perspective on investing in digital
technologies, drawing on the resource-based view of the firm theory (RBV),
motivating you to reflect on how businesses build sustainable competitive
advantages in a digital world.
This course emphasises the role of the internet, in many ways the most
important digital infrastructure of our times. Indeed, all the other digital
infrastructures discussed in this guide depend on the internet for data
transmission. The WWW has been created to make it possible to find and
share content on the internet. Cloud computing uses the internet for the
transmission of data to support all manner of digital services. Social media
too is a form of digital infrastructure on the top of the internet. Finally,
mobile technology has become so pervasive because of its connection with
the internet. People place mobile devices at the centre of their professional
and personal activities because of the connection with the internet in
general (smartphones) and particularly because mobile devices have
become key channels to access information, access services and social
media. These five digital infrastructures (internet, web, cloud, social,
mobile) are deeply connected to each other and the development of any of
them is going to affect the others and related digital products and, services
and business models.
Lastly, we have discussed the importance of governance and regulation of
digital infrastructures and how different bodies and interests are drawn in
to maintaining their coherence as they evolve. Some of this is in the realm
of governments and laws, but much lies beyond in independent institutions
such as the IETF or W3C. Without appropriate governance and regulation
infrastructures cannot thrive and grow. The evidence we see is that just as
the technical aspects of any infrastructure are varied and heterogeneous,
so too are the governance aspects. Governance is a mosaic of bodies
and responsibilities, rather than being achieved by a single formal set of
structures. We have also discussed how establishing common governance
rules and regulations is a very political endeavour and. In particular, we
have discussed the case of net neutrality, to exemplify the difficulties to
build consensus on controversial issues.
232
20.2.2 The new frontiers of digital innovation

This section explores some aspects of where and how digital technologies
are evolving in the next decade or so. However, we will not make futuristic
projections, but stay grounded in digital technologies that are already
working and which promise bigger impacts in the near future.
There is probably no controversy in saying that communications, work
and social activities, entertainment, education, getting information and
buying/selling have migrated in some significant degree to the digital
sphere. Depending on where you live this will be to differing degrees.
The more digital infrastructures become available, the more people join
them to get the benefits, as discussed throughout this subject guide. The
pervasiveness of digital technologies changes the way we behave and
the way we understand ourselves and our behaviour. From a macro-
perspective, economists and strategists have been using the term the fouth
Industrial Revolution to express the impact of advanced technologies in
the economy, particularly digital technologies (IoT, big data analytics, AI,
quantum computing, robotics, etc.).
The more we are online interacting for different reasons, the more data is
generated about us and our activities. In the past, getting such data on the
activities of large populations would be very costly in terms of time and
money. Today, a great part of our communication patterns, location and
travel, financial transactions and purchases are observed and recorded. Our
digital interactions leave traces in the digital sphere. Every time we navigate
the internet, there is one or more algorithm registering our behaviour, such
as Google, Facebook, Amazon and WeChat or any other company which has
placed cookies in our computing interfaces. Every time we send an email,
we leave a trace to be analysed by the service provider or the company.
Every time we collaborate or interact online, we leave a trace about our
contribution to a project and our way of thinking, as well as our mistakes.
Every time we use our smartphones, the service provider knows who we are
talking to, what information we seek and where we are when we want it.
This phenomenon of automatic generation of digital data which is stored
in databases has been called big data. Generating data, however, is not
very useful per se. Big data is a resource to be used by professionals
specialising in analytics. Big data analytics makes use of the mountains
of data we generate. Through analytics, big datasets can reveal new
information that we could not know without data on this scale and at this
level of granularity. For instance, we can know more about patterns of
behaviour, unknown correlations and influences, customer preferences,
public opinion trends, productivity and levels of collaboration, to cite
a few examples. When big data is available, those companies that can
exploit this resource to gain knowledge of the market gain competitive
advantages in relation to those which are not. There are thus direct
economic benefits from big data analytics for businesses.
The potential use of big data has increased in the last years with the
gradual advance of another digital infrastructure, which has been called
the IoT (Chapter 14). The IoT is about adding digital sensors with network
connectivity to objects, thus allowing the communication of these objects
between themselves and with the internet in a broad sense. For instance,
with this infrastructure in place, people can manage their home devices
through the internet. They can tell the heating system or air conditioner to
start when they are on their way home. Other possibilities are to turn on the
oven to warm food, or a food dispenser to release food to the pets. Such IoT
infrastructure will in time change the way we interact with our homes.
233
On a broader scale, the IoT can create, for instance, smart electrical
grids, intelligent transport systems and smart cities, changing the way
governments and private companies manage urban spaces. Imagine an
electrical grid in which all devices are connected to the same network
and the grid itself can decide which time is the best to do some tasks
at home – for instance, washing clothes or heating water – when there
is excess of energy in the system. Or an intelligent transport system
which automatically directs passengers, cars and buses to the fastest
routes in accordance with the information received from other vehicles
in the network, avoiding traffic. Or an IoT infrastructure helping a city
to manage in a more intelligent way schools, hospitals, services, power
plants, water supply and waste management. These activities, individually
and integrated together, can be better managed using the information
from diverse sensors. The result could be reduction of costs and waste and
the improvement of the quality of services, through better productivity,
less pollution and greenhouse gasses.
IoT has other potential applications, for instance in self-driving vehicles
and electric powered – cars, buses and trucks. Vehicles can quite easily
understand maps to go from point A to point B without a human driver
and petrol engines can be changed to electric motors. But how to do
it safely and reliably? For that we need to understand many material
elements in the external context – such as other cars, bicycles, buildings,
people and animals, weather, charging points, journey plans etc. This
needs a lot of data to be collected and then processed – probably remotely
by multiple dedicated services. The development of electric vehicles
and self-driving cars calls attention to another familiar phenomenon: all
infrastructures evolve and grow, drawing on an installed base. The new
digital transport infrastructure emerging builds on and extends existing
transport infrastructures (roads, gas stations, maps) and blends this
with mobile digital communications. We can foresee a new transport
infrastructure built on top of digital infrastructures – including smarted
electric vehicles, public transport, trains, bicycles, ships and aeroplanes.
A final development which is going to deeply affect the future of
computing and its implications for businesses is AI (Chapter 15). Our view
is that the day when computers will be able to match the human brain
(what some call the Artificial General Intelligence) is far off. However,
there are some narrower and specific areas in which computers are already
better than humans, for instance, managing massive volumes of data and
finding unknown patterns (associations between data points). In addition,
computer scientists have learnt that one successful way to build AI is
through mechanisms which mimic the process of human learning. Instead
of trying to develop a software programme which ‘knows’ everything,
better to have a programme which can learn from past experience – e.g.
from databases and from interaction with humans. This is known as the
machine learning (ML) approach. To some degree, using ML, programs
can interpret natural language (spoken or written) and answer questions
on relevant topics, looking for a perfect match or anything meaningful,
within a range of probabilities, in the same way human beings do.
By combining big data with ML techniques AI today has become a very
viable digital service. It is not about ‘knowing’ everything, but learning how
to learn in specific domains and making a better estimation of probabilities
to match questions to answers. Even better, having developed general
models of learning, they can be applied to the learning of other topics
that has a suitable body of data to be analysed. Once an AI algorithm is
trained for a task, the cost of reproducing that algorithm (AI as a Service)
234
is close to zero (Chapter 9). If an AI algorithm is more precise than humans

in driving cars (still a case of ‘if’ today), it may be possible to substitute all
drivers in the world with a single algorithm, but only if we reach a high
level of confidence in it. A similar logic may apply in other areas as for
example the diagnosis of diseases from scans and X-rays, decision making
about investments, or even the teaching of some subjects.
Increased roles for AI require a broad discussion in society to understand
the advantages and risks of its use and production. AI using ML depends
on access to a large volume of data (big data) that captures some causal
relationships. Those with access to relevant datasets have competitive
advantages in relation to others. We also face the problem that the data
we use to train an AI algorithm may be significantly biased leading to
systemically biased AI services. If we are pessimistic, we might see AI as
creating an increased gap between the haves and have-nots in the domain
of digital technologies; those with access data and infrastructures to build
and use AI resources and those without.
Advances in the practical application of AI has made possible the
identification of patterns of behaviour, language, voice and images that
we could not imagine a few decades ago. These applications have enabled
positive enhancements in products and services. For instance, you can use
your voice or face image to open your mobile and have access to financial
services through applications. It is safer to use biometric information than
passwords. But it is also the case that the same technology can be used to
identify and control people. The more digitalised our lives become, the
more information is gathered about our characteristics, behaviour and
beliefs, the higher the risk that such information can be used against us.
And to make the digital game even more interesting, we also need to
add a small introduction to quantum computing. Instead of relying on
the traditional computing power of binary information (each element in
the computing process is either a one or a zero, which are combined for
representing all data you have digitalised), the quantum computer process
information considers the quantum states of the system. Without going very
deep into the technical aspects of quantum computing (this is not the place
to discuss such a complex technology), the expectations are that quantum
computers are going to be able to tackle much more complex problems
and much more quickly. Indeed, problems that today we cannot address
in a feasible amount of time. In this case, speed makes all the difference.
When quantum computers are fully operational, as per the theory, their
speed to calculate things will be so fast that people and companies using
the current technology will not be able to be competitive anymore (at
least in the processing of data). We can imagine the significant advances
in AI algorithms, for instance, using quantum computing. Companies like
Google and IBM are focusing on developing commercial applications for the
quantum technology. We are not yet there and me way never get there, but
it is good to pay attention to this possible development, which may change
radically the way we understand computing.
Activity 20.1: Big data analytics for digital marketing

Marketing professionals have been very quick to use big data analytics to get better
results in their campaigns and to build their brands more effectively. Research three
examples in which big data analytics have helped companies to better reach their
marketing goals. Compare the three examples and see the differences and similarities in
strategies. You can also look for professionals which are specialists in big data analytics
for marketing. There are plenty of examples on YouTube.
235
Activity 20.2: Baxter robot

The advances in robotics are changing our understanding about what robots can do.
An interesting example to see is the robot called Baxter. Check on Wikipedia (https://
en.wikipedia.org/wiki/Baxter_(robot)) what Baxter can do. Now decide how Baxter can
be of help to corporations. Also check the software approach which makes Baxter an
interesting example of intelligent robots. You can also find plenty of videos on YouTube
about Baxter.
Activity 20.3: The surveillance of citizens and users

Have you ever thought about how much data governments and companies have about
you and your behaviour? Make two lists of relevant information private companies and
governments (in your country and beyond) have about you. Now reflect upon how these
companies and governments may use this information. Think about two scenarios: one
in which the information is used for your benefit and another in which the information is
used in some way against your interests.
20.2.3Digital infrastructure innovation and the future of

businesses
In this final section, we explore a few implications and concerns deriving
from our broad discussion of digital infrastructures. There are two aspects
we emphasise: one is about the survival of businesses and economies and
the other is about the need to change education systems and the necessary
skills for this new era of pervasive digital infrastructures. This section
is not for giving answers, but to call attention to problems which will
require more care in the future. You can read more about the theme in
Brynjolfsson and McAfee (2011, 2014).
From the perspective of businesses and economies, there are not many
alternatives to embracing the adoption of digital infrastructures. Naturally,
there are always some niches which can survive or even thrive with less
digital technology. But the great majority of businesses are dependent on
digital infrastructures and will become more so. Having more connections
to buyers and seller, more access to information and AI services and
capacity for processing information at a higher speed, gives a competitive
edge to individual companies and national economies. As discussed in this
guide, digital infrastructures help drive business innovation, potentially
bringing more creativity, more productivity, better business models and
new digital products and services.
From the perspective of educational systems the scale of the challenge
is no less dramatic. The problem here is seen by some experts as the
mismatch between traditional education systems, which were preparing
people for economies before the emergence of digital infrastructures
and the current needs. Seen narrowly, many industries face difficulties
to recruit engineers, computer scientists, data analysists, programmers
and business professionals who are able to lead the innovation of digital
infrastructures. The situation becomes more serious if we think that AI
is going to substitute many professional roles partially or fully. Those
affected by AI solutions cut across professions and roles – including all
kinds of knowledge workers such as bankers, doctors, teachers, and
lawyers to name a few. Many people will need to reinvent themselves in
some degree and orient towards areas where AI solutions do not fit.
There is no easy way out of this situation in which educational institutions
cannot change as fast as the needs of economies and societies require. One
possible route is to change our education systems to be more focused on
236
the process of ‘learning how to learn’. We have tried in this subject guide to
show you some steps in this direction. Many of our activities are about you
doing your own research to find your own answers to problems. Instead
of trying to present ‘all that you need to know’ about a topic, this subject
guide tries to tell you a story which fosters your interest in looking for more
information through the recommended readings and through your own
research. In doing the activities in this guide, you become more capable of
taking the process of learning a complex topic in your own hands.
Activity 20.4: Online free courses on digital infrastructures and digital

business innovation
There are online some very interesting free courses on the topics covered in this subject
guide. Choose three topics of your preference and go online to find free courses on these
topics. Most of these courses are MOOCs (Massive Online Open Courses). As a tip, start
researching Coursera, an online course interface (https://www.coursera.org/). There are
other Moocs interfaces. Explore them.

This chapter focused on two aspects. First, it summarised key topics and
theories presented in the previous chapters. The objective of this summary
is to help you to have an overview of the course, being able to see how the
parts come together. Second, this chapter explored a few topics which are
still in development, but which are likely to affect deeply the development
of digital infrastructures and business in the next few decades. You should
end the course with two perspectives on digital infrastructures. The first is
about the current situation, which already is there for us to explore, such
as the internet and the WWW, cloud computing, mobile technologies and
social media. The second is about understanding that digital innovation is
evolving quickly, changing our understanding of digital infrastructures. We
need to be prepared for the challenges these innovations are going to bring
to our lives.

• explain how the parts (chapters) are combined into the whole course,
IS1181 Digital infrastructure for business
• describe current innovations which are affecting digital infrastructures
• explain how innovations in digital infrastructures are challenging
organisations, individuals and societies
• outline individual and institutional strategies for continuously learning
to adapt to changes in a context of constant innovation in digital
infrastructures.
237

1. Develop a short description of three key theoretical frameworks
discussed in this course and apply it to one or more topics related to
the future of digital infrastructures.
2. How can AI and process automation be delivered ‘as a service’ using
digital infrastructures?
3. What do you believe are the key challenges you are going to face
in your professional career in the next decade because of changes
consequent on digital infrastructures? What is your strategy to cope
with these challenges
238

Digital Infrastructure - Subject Guide

Uploaded by

Copyright:

Available Formats

Digital Infrastructure - Subject Guide

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Digital Infrastructure - Subject Guide

Uploaded by

Copyright:

Available Formats

Undergraduate study in Economics,

Management, Finance and the Social Sciences

Chapter 1: Introduction to the course..................................................................... 1

Chapter 7: Core ideas for evolving infrastructures I: installed base

Chapter 13: Social media as an innovation in infrastructure.............................. 145

Chapter 20: Reflections on the course: the future of digital infrastructures

Chapter 1: Introduction to the course

1.1 Route map to the guide

‘The encoding of analogue information into a digital format and

In this spirit of ‘consequential change’, we will consider in this course

1.2 Introduction to the subject area of digital

In this way the most significant and pervasive digital infrastructure,

neutrality and national borders. These issues illustrate how governance

1.4 Aims of the course

1.5 Learning outcomes for the course

1.6 Overview of learning resources

1.6.1 The subject guide

subsequent lectures and chapters, reducing your chances of getting the

1.6.2 Essential reading

information systems. We can also usually assume that the organisation

1.6.3 Further reading

• Past examination papers: We provide up to three years of past

Making use of the Online Library

Chapter 2: Infrastructures: conceptual

2.1.1 Aims of the chapter

2.1.2 Learning outcomes

2.1.3 Essential reading

2.1.4 Further reading

Social media: https://en.wikipedia.org/wiki/Social_media

2.1.5 References cited

2.1.6 Synopsis of chapter content

2.2 Chapter content

Third, these definitions give examples of things that are considered to

2.2.2 Examples of infrastructures: electricity and rail networks

migrate from lots of different local arrangements to a large scale and

2.2.3 Infrastructures, production and organisation

because of its capacity of bringing value to organisations and consumers

2.2.4 The emergence of digital infrastructures

As outlined in Chapter 3 below, the internet was initially designed to

• Shared: a digital infrastructure is shared by the members of a community who use

2.2.5 Digital infrastructure and business

Activity 2.2: Key ideas behind the concept of digital infrastructure

2.2.6 Innovation and digital infrastructure

Activity 2.3: Key ideas behind the concept of digital infrastructure

2.2.7 Examples of digital infrastructures

The world wide web (WWW)

Activity 2.4: The time for cloud computing to spread

contents, including texts, photos, images and videos, in addition to voice.

Activity 2.5: The relevance of social media

2.2.8 New products, services and business models based on digital

Activity 2.6: The internet as infrastructure

Activity 2.7: How have digital infrastructures enabled a business such as

2.3 Overview of chapter

2.4 Reminder of learning outcomes

2.5 Test your knowledge and understanding

Chapter 3: History of the internet

3.1.1 Aims of the chapter

3.1.2 Learning outcomes

3.1.3 Essential reading

1.2 Introduction to the subject area of digital

2.2.8 New products, services and business models based on digital

3.2.4 Internet evolution: from a scientific endeavour to a universal

4.2.2 Use packet switching to share bandwidth and build in